UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Degradation of diethanolamine solutions Kennard, Malcolm L. 1983-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
UBC_1983_A1 K46.pdf [ 11.97MB ]
Metadata
JSON: 1.0058802.json
JSON-LD: 1.0058802+ld.json
RDF/XML (Pretty): 1.0058802.xml
RDF/JSON: 1.0058802+rdf.json
Turtle: 1.0058802+rdf-turtle.txt
N-Triples: 1.0058802+rdf-ntriples.txt
Original Record: 1.0058802 +original-record.json
Full Text
1.0058802.txt
Citation
1.0058802.ris

Full Text

DEGRADATION OF DIETHANOLAMINE SOLUTIONS by MALCOLM L. KENNARD B.Sc, University of Nottingham, England, 1974 M.A.Sc., University of British Columbia, 1978 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Chemical Engineering We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January 1983 © Malcolm L. Kennard, 1983 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date DE-6 (.3/81) ABSTRACT Raw natural gas contains acid gases such as H2S and C02 which must be removed before the gas can be sold. The removal of these gases is called "sweetening" and the use of Diethanolamine (DEA) as a solvent has become widely accepted by industry. The process is simply based on the absorption and desorption of the acid gases in aqueous DEA. Side reactions can occur when DEA reacts with the C02 to produce degradation compounds. This degradation causes a loss in valuable DEA and an increase in plant operating costs. The reaction between DEA and C02 was studied experimentally, using a 600 ml stirred autoclave, to determine the effect of temperature, DEA concentration, and reaction pressure. Degraded DEA samples were analysed using gas chromatography. A fast, simple, and reliable technique was developed to analyse degraded DEA samples, which was ideally suited to plant use. . Over 12 degradation compounds were detected in the degraded DEA solutions using gas chromatography and mass spectroscopy. Degradation mechanisms are proposed for the production of the various compounds. It was found that the degradation of DEA was very sensitive to temperature, DEA concentration, and C02 solubility of less than 0.2 g C02/g DEA. To study the effect of C02 solubility, which is a function of reaction pressure, simple solubility experiments were performed to cover the range of 100-200°C, 413.7-4137 kPa (60-600 psi) partial pressure of C02 and DEA concentration of 10, 20, and 30 wt % DEA. It was found that the reaction between DEA and C02 was extremely complex consisting of a mixture of equilibria, parallel, series, and ionic reactions. However, the overall degradation of DEA could be simply described by a pseudo first order reaction. The main degradation products were HEOD, THEED, and BHEP. It was concluded that C02 acted as a catalyst being neither consumed nor produced during the degradation of DEA to THEED and BHEP. HEOD was produced from DEA and C02, but was found to be unstable and could be converted back to DEA or react to form THEED and BHEP. The following simple kinetic model was developed to predict the degradation of DEA and the production of the major degradation compounds:-DEA BHEP The model covered the ranges of DEA concentration 0-100 wt % DEA, 90-175°C, and C02 solubilities greater than 0.2 g C02/g DEA. Attempts were made to purify degraded DEA solutions. It has been claimed that activated carbon filters are useful in removing degradation compounds. However, tests with activated carbon proved it to be inca pable of removing any of the major degradation compounds. See Nomenclature TABLE OF CONTENTS ABSTRACT ,ii LIST OF TABLES ix LIST OF FIGURES xACKNOWLEDGEMENTS v Chapter 1 INTRODUCTION 1 1.1 Objectives of the present study 4 2 LITERATURE REVIEW 5 2.1 Absorption of C02 in aqueous DEA solutions 5 2.2 DEA degradation 9 2.3 Analysis of DEA and its degradation products ... 13 3 DEVELOPMENT OF THE ANALYTICAL TECHNIQUE 17 3.1 Gas chromatographic technique 18 3.1.1 Evaluation of the Tenax G.C. column 19 3.1.2 Operating conditions 13.2 Analytical procedure and performance 21 3.2.1 Column performance 23.3 G.C. calibration 23 3.4 Maintenance of chromatographic equipment 23.5 Advantages of the analytical technique 8 3.6 Errors 29 3.6.1 Accuracy 23.7 Units of DEA concentration 30 4 SYNTHESIS OF SELECT DEGRADATION COMPOUNDS FOR CALIBRATION OF THE GAS CHROMATOGRAPH 31 4.1 Synthesis of HEOD 34.2 THEED synthesis 4 5 IDENTIFICATION OF DEGRADATION COMPOUNDS 38 5.1 Identification using the gas chromatograph 38 5.2 Identification using a GC/MS 39 5.3 Identified degradation compounds 40 iv 6 EXPERIMENTAL EQUIPMENT AND PROCEDURE FOR THE CONTROLLED DEGRADATION OF DEA 45 6.1 600 ml. autoclave 46.2 Loading the autoclave 7 6.3 Sampling 49 6.4 Analysis of the liquid samples 46.5 Analysis of the gas phase 50 6.6 Experimental procedure6.7 Maintenance and performance 2 6.8 Sources of errors 57 PRELIMINARY EXPERIMENTS AND DEVELOPMENT OF THE EXPERIMENTAL PROGRAMME 54 7.1 Use of high temperatures for the degradation runs 57.1.1 Temperature comparisons 54 7.1.2 Comparison with industrial samples 56 7.1.3 Thermal degradation7.1.4 Justification for the use of elevated temperatures 56 7.2 Effect of metal surfaces 58 7.3 Effect of stirrer speed and reactant volume 58 7.4 Reproducibility 9 7.4.1 Samples 57.4.2 Runs7.5 Experimental programme , 60 8 RESULTS AND DISCUSSION OF EXPERIMENTS DESIGNED TO STUDY THE KINETICS OF THE DEGRADATION REACTION 62 8.1 Effect of temperature 64 8.1.1 Degradation products 72 8.1.1.1 HEOD 78.1.1.2 THEED8.1.1.3 BHEP 9 8.2 Effect of initial DEA concentration 78.3 Effect of pressure 84 9 EXPERIMENTS DESIGNED TO ELUCIDATE THE DEGRADATION MECHANISM 95 9.1 Effect of pH 99.2 Effect of bicarbonate and carbonate ions 98 9.2.1 C02 solubility data 99.2.2 Runs using HC03~ and C03 instead of C02 103 9.3 Effect of water 108 9.4 Thermal degradation 115 9.5 C02 solubility studies 120 v 10 EXPERIMENTS DESIGNED TO STUDY THE BEHAVIOUR OF THE MAJOR DEGRADATION COMPOUNDS AND IMPURITIES IN THE DEA FEED 125 10.1 Long term run 1210.2 BHEP runs 8 10.3 HEOD runs 129 10.4 THEED runs 137 10.5 Experiments to study the effect of impurities in the DEA feed 142 11 DEVELOPMENT OF A MECHANISM FOR DEA DEGRADATION 145 11.1 Formation and reactions of the major degradation compounds 146 11.1.1 Formation of HEOD 1411.1.2 Behaviour of HEOD under reaction conditions 147 11.1.2.1 Proof that BHEP is not produced directly from HEOD .. 147 11.1.2.2 Equilibrium between HEOD and DEA carbamate 150 11.1.2.3 Proof that THEED is not pro duced directly from HEOD 153 11.1.3 Proposed model for the production and reactions of HEOD 1511.1.4 Effect of temperature on the production of HEOD 157 11.1.5 Reaction of HEOD and DEA under N2 158 11.1.6 Formation of THEED 1511.1.7 Formation of BHEP 160 11.2 Discussion of the degradation routes 1611.2.1 Ionic route 1611.2.2 Molecular route 162 11.2.2.1 Molecular runs 1611.2.3 Thermal route 163 11.3 Discussion of anomalous experimental observations 164 11.3.1 The relationship between initial kj)EA and DEA concentration 164 11.3.2 Arrhenius plot 1611.3.3 Log [DEA] versus time plots 166 11.3.4 Explanation of the effect of pH 168 11.4 The formation of minor degradation compounds ... 169 11.4.1 MEA degradation 1611.4.2 Reaction between MEA and DEA 172 11.4.3 Minor degradation compounds produced from DEA 174 11.4.4 The reaction between DEA and TEA 175 11.5 Summary 1711.5.1 Conclusions of the degradation experiments 175 vi 12 KINETIC STUDIES 180 12.1 Development of a kinetic model 180 12.1.2 Simplified degradation mechanism 181 12.2 Theory 183 12.3 Calculation of the k values 184 12.3.1 Method (A)—The plot of [THEED] vs. t goes through a maximum 184 12.3.2 Method (B)—The plot of [THEED] vs. t does not go through a maximum 185 12.4 Comparison of the experimental results with the predictions of the model 188 12.5 Application of the model13 PURIFICATION. OF DEGRADED DEA SOLUTIONS 194 13.1 Use of activated carbon 1913.2 Use of solvents 195 13.3 Removal of BHEP 8 13.4 Removal of HEOD13.5 Conclusion 199 14 CONCLUSIONS AND RECOMMENDATIONS 200 14.1 Practical implications of the present study 202 NOMENCLATURE 205 REFERENCES 7 Appendices A Sources of equipment and chemicals 212 B Experimental results for the degradation of DEA by C02 214 C The solubility of C02 in DEA solutions at high temperature and pressure 249 C.l Experimental methodC.2 Calculation of C02 solubility 252 C.3 Example 253 C.4 Results 4 D Derivation of the kinetic model 258 E Comparison between the experimental results and the prediction of the kinetic model 261 vii F Mass spectra of DEA and related degradation compounds 277 F.l Mass spectra of DEA and its degradation compoundsF.2 Mass spectra of minor degradation compounds 277 F.3- Mass spectra of impurities in the DEA feed 27F.4 Mass spectra of associated compounds 277 F.5 Summary 27viii LIST OF TABLES Table 2.1 Equations governing C02 absorption in aqueous DEA solution 6 3.1 Analytical equipment and operating conditions for G.C. analysis 20 5.1 Compounds found in degraded DEA solutions 44 7.1 Comparison of reproducibility of degradation runs .... 59 7.2 Summary of experiments performed with C02 at 4137 kPa (600 psi) 60 8.1 Initial DEA concentration and initial DEA degradation rates at 205°C 68 8.2 Initial DEA concentration and initial DEA degradation rates at 175°C8.3 Initial DEA concentration and initial DEA degradation rates at 150°C 70 9.1 Comparison of C02 solubilities in DEA solutions 99 9.2 Summary of HC03_ and C03— runs 103 9.3 Comparison of k^p. values for runs conducted with KHC03 and C02 , 108 9.4 Summary of the molecular runs 109 9.5 Comparison of kn£A ^or m°lecular> ionic, and standard runs 114 9.6 Summary of thermal runs 115 9.7 Comparison of k^^ for thermal and standard runs 115 9.8 Overall k^EA as a function of C02 solubility for degradation runs of 30 wt % DEA at 195°C 120 9.9 C02 solubility as a function of temperature in a 30 wt % DEA solution under a total pressure of 4137 kPa (600 psi) 122 9.10 C02 solubility as a function of DEA concentration for solutions at 205°C under a total pressure of 4137 kPa (600 psi) 1210.1 Summary of runs to study the behaviour of the major degradation compounds 128 10.2 for reactions under CO, and N2 142 10.3 Experimental conditions of runs performed to determine the effects of DEA feed impurities 142 11.1 Principal DEA degradation routes under various conditions 176 13.1 Effect of various solvents on degraded DEA solutions .. 198 ix B.l Run 1 215 4- 4 B. 73 Run 73 248 C. l Vapour pressure of DEA solutions as a function of temperature 250 C.2 Solubility of C02 in 30 wt % DEA 255 C.3 Solubility of C02 in 20 wt % DEA 6 C.4 Solubility of C02 in 10 wt % DEA 257 E.l Comparisons between experimental and predicted concentrations 262 4- 4-E. 15 Comparisons between experimental and predicted concentrations 276 F. l Molecular formula and major peaks of mass spectra of compounds studied 293 x LIST OF FIGURES Figure 1.1 Flow sheet of the basic amine process 2 2.1 Shuttle mechanism for C02 absorption into aqueous amine solutions 8 3.1 Typical chromatogram for MEA, DEA, and TEA 22 3.2 Typical chromatogram of a degraded DEA solution 22 3.3 Calibration plot for DEA 24 3.4 Calibration plot for HEOD 5 3.5 Calibration plot for BHEP 6 3.6 Calibration plot for THEED 27 4.1 Chromatogram of substituted ethylenediamines ... 36 5.1 Chromatogram showing peak of unknown compound 41 5.2 Mass spectrum of HEOD 42 5.3 Mass spectrum of unknown compound 45.4 Typical chromatogram of a degraded DEA solution 43 6.1 Sketch of 600 ml. autoclave 46 6.2 Sketch of the autoclave loading system 48 7.1 Chromatogram of a 30 wt % DEA solution degraded at 205°C under 4137 kPa C02 (time—1 hr.) 55 7.2 Chromatogram of a 30 wt % DEA solution degraded at 120°C under 4137 kPa C02 (time—150 hr.) 55 7.3 Chromatogram of a degraded DEA solution from a gas sweetening unit operated at Aquitaine Canada Ltd 57 7.4 Chromatogram of a degraded DEA solution from a gas sweetening unit operated by Hudson's Bay Oil and Gas Co. Ltd 58.1 Typical chromatograms of a 30 wt % DEA solution degraded at 205°C under 4137 kPa 63 8.2 DEA concentration as a function of time and temperature (4137 kPa C02 , 205-162°C) ' 65 8.3 DEA concentration as a function of time and temperature (4137 kPa C02 , 150-90°C) 66 8.4 Initial degradation rate as a function of initial DEA concentration and temperature 9 8.5 Arrhenius plot for a 30 wt % DEA solution degraded with C02 at 4137 kPa 71 8.6 HEOD concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 205-162°C) ... 73 8.7 HEOD concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 150-90°C) 74 8.8 THEED concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 205-162°C) ... 75 xi ) 8.9 THEED concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 150-90°C) 76 8.10 BHEP concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 206-162°C) ... 77 8.11 BHEP concentration as a function of time and temperature (30 wt % DEA, 4137 kPa C02, 150- 90°C) ... 78 8.12 DEA concentration as a function of time and initial DEA concentration (4137 kPa C02 , 205°C) 80 8.13 DEA concentration as a function of time and initial DEA concentration (4137 kPa C02 , 175°C) 81 8.14 DEA concentration as a function of time and initial DEA concentration (4137 kPa C02 , 150°C) 82 8.15 knj?A as a function of initial DEA concentration and temperature (4137 kPa C02) 83 8.16 Arrhenius plot for varying DEA solution strengths degraded with C02 at 4137 kPa 5 8.17 HEOD concentration as a function of time and initial DEA concentration (4137 kPa C02 , 205°C) 86 8.18 THEED concentration as a function of time and initial DEA concentration (4137 kPa C02 , 205°C) 87 8.19 BHEP concentration as a function of time and initial DEA concentration (4137 kPa C02 , 205°C) 88 8.20 DEA concentration as a function of time and C02 pressure (30 wt % DEA, 195°C) 90 8.21 k^,,. as a function of reaction pressure (30 wt % DEA, DtA 195°C) 91 8.22 HEOD concentration as a function of time and C02 pressure (30 wt % DEA, 195°C) 92 8.23 THEED concentration as a function of time and C02 pressure (30 wt % DEA, 195°C) 3 8.24 BHEP concentration as a function of time and C02 pressure (30 wt % DEA, 195°C) 94 9.1 DEA concentration as a function of time and solution pH (30 wt % DEA, 4137 kPa C02 , 205°C) 96 9.2 Solubility of C02 in 30 wt % DEA 100 9.3 Solubility of C02 in 20 wt % DEA 1 9.4 Solubility of C02 in 10 wt % DEA 102 9.5 DEA concentration as a function of time and temperature (using KHC03, 30 wt % DEA, 4137 kPa N2) .. 104 9.6 HEOD concentration as a function of time and temperature (using KHC03, 30 wt % DEA, 4137 kPa N2) .. 105 9.7 THEED concentration as a function of time and temperature (using KHC03, 30 wt % DEA, 4137 kPa N2) .. 106 9.8 BHEP concentration as a function of time and temperature (using KHC03, 30 wt % DEA, 4137 kPa N2) .. 107 9.9 DEA concentration as a function of time and initial DEA concentration (no water present, 4137 kPa C02, 205°C) 110 9.10 HEOD concentration as a function of time and initial DEA concentration (no water present, 4137 kPa C02, 205°C) Ilxii 9.11 THEED concentration as a function of time and initial DEA concentration (no water present, 4137 kPa C02, 205°C) 112 9.12 BHEP concentration as a function of time and initial DEA concentration (no water present, 4137 kPa C02, 205°C) 113 9.13 DEA concentration as a function of time and temperature (no C02 present, 30 wt % DEA, 4137 kPa N2) 116 9.14 THEED concentration as a function of time and tempera ture (no C02 present, 30 wt % DEA, 4137 kPa N2) 118 9.15 BHEP concentration as a function of time and tempera ture (no C02 present, 30 wt % DEA, 4137 kPa N2) 119 9.16 k„. as a function of CO, concentration (30 wt % DEA 2 DEA, 195°C) 121 9.17 DEA concentration as a function of time and C02 pressure (30 wt % DEA, 205°C) 124 10.1 Typical plots of concentration as a function of time for DEA and its degradation products 126 10.2 Concentration of DEA, HEOD, THEED, and BHEP as a function of time (30 wt % DEA, 4137 kPa C02, 205°C) .. 127 10.3 DEA concentration as a function of time and degradation products (30 wt % DEA, 4137 kPa C02 , 205°C) 130 10.4 Concentration of DEA, HEOD, THEED, and BHEP as a function of time (reactants—DEA and HEOD, 4137 kPa C02 , 205°C) 132 10.5 DEA concentration as a function of time (reactants— degraded DEA solution, 4137 kPa C02 or N2, 175°C) 133 10.6 HEOD concentration as a function of time (reactants— degraded DEA solution, 4137 kPa C02 or N2, 175°C) 134 10.7 THEED concentration as a function of time (reactants— degraded DEA solution, 4137 kPa C02 or N2, 175°C) 135 10.8 BHEP concentration as a function of time (reactants— degraded DEA solution, 4137 kPa C02 or N2, 175°C) 136 10.9 Concentration of DEA, THEED, and BHEP as a function of time (reactants—DEA and THEED, 4137 kPa N2, 205°C) 138 10.10 Concentration of DEA, HEOD, THEED, and BHEP as a function of time (reactants—DEA and THEED, 4137 kPa C02 , 205°C) 140 10.11 logfTHEED] as a function of time (reactants—DEA and THEED, 4137 kPa C02 or N2 , 205°C) 141 10.12 Typical chromatogram of a degraded solution of DEA and MEA 144 11.1 Qualitative plots of concentration versus time, showing the possible relationships between HEOD and BHEP 148 11.2 Sketch of the concentration of HEOD and BHEP as a function of time for run 65 1411.3 Sketch of the concentration of HEOD and BHEP as a function of time for runs 8 and 10 151 11.4 Sketch of the concentration of HEOD and THEED as a function of time for runs ,1 to 12 154 11.5 Schematic diagram of degradation of DEA 157 xiii 11.6 Sketch of k^^ as a function of intial DEA concentration 165 11.7 Sketch of the Arrhenius plots for ionic, molecular, and standard runs 7 11.8 Schematic diagram showing the possible routes for the degradation of DEA 179 12.1 Typical plot of [THEED] versus time 184 12.2 Comparison between the experimental and theoretical values of DEA, HEOD, THEED, and BHEP concentrations as a function of time (run 6) 186 12.3 Comparison between the experimental and theoretical values of DEA, HEOD, THEED, and BHEP concentrations as a function of time (run 23) 187 12.4 Arrhenius plots for ki at various initial DEA concentrations (DEA HEOD) 189 12.5 Arrhenius plots for k2 at various initial DEA concentrations (DEA -^-*THEED) 190 12.6 Arrhenius plot for k3 at various initial DEA concentrations (THEED -^-BHEP) 191 12.7 Sketch of Arrhenius plots for ki and k2 at various initial DEA concentrations 193 13.1 Typical chromatograms of partially degraded DEA solutions taken upstream and downstream of an activated carbon filter located in a large gas plant 196 13.2 Typical chromatograms of partially degraded DEA solutions under laboratory conditions; before and after contact with activated carbon 197 C.l DEA solution vapour pressure as a function of temperature and DEA concentration 251 F.l Mass spectrum of DEA 278 F.2 Mass spectrum of BHEP 9 F.3 Mass spectrum of HEOD 280 F.4 Mass spectrum of THEED 1 F.5 Mass spectrum of BHEED 2 F.6 Mass spectrum of HEED 283 F.7 Mass spectrum of HEI 4 F.8 Mass spectrum of HEM 5 F.9 Mass spectrum of HEP 286 F.10 Mass spectrum of OZD 7 F.ll Mass spectrum of TEHEED 288 F.12 Mass spectrum of MEA 9 F.13 Mass spectrum of TEAF.14 Mass spectrum of BHG 290 F.15 Mass spectrum of DEEA 1 F.16 Mass spectrum of MDEA 2 xiv ACKNOWLEDGEMENTS The author wishes to thank the following:-- Dr. Axel Meisen for his supervision and guidance throughout the course of this work; - Nina Thurston for her excellent typing; - Patricia Parsons who drew the figures. The financial assistance provided by the Canadian Gas Processors Association for this work is gratefully acknowledged. xv CHAPTER 1 INTRODUCTION Raw natural gas leaving the geological formation frequently contains undesirable compounds such as hydrogen sulphide, carbon dioxide, and water. The acid gases and the water must be removed before the natural gas can be sold in order to minimize pipe line corrosion, pumping costs, health hazards, and pollution when the gas is finally burnt or converted. Purification, or 'sweetening', of large volumes of 'sour' natural gas requires a process which is efficient and inexpensive. Further, in order for a purification process to be efficient, it must be regenerative, simple, and relatively trouble free. The use of aqueous diethanolamine (DEA) has emerged as the best approach to the sweetening of sour natural gas. The process is based on the reaction of a weak base (DEA) with a weak acid (H2S or C02) to give a water soluble salt. The reactions may be summarized as follows:-(H0-C2HJ2 NH + H2S c=± (H0-C2H1|)2 NH2+ HS" [1.1] (H0-C2H4)2 NH + C02 + H20 ^ (H0"C2H4)2 NH2+ HC03~ [1.2] These reactions are reversible and the equilibria may be shifted by regulating the temperature. The basic industrial process (see Fig. 1.1) consists essentially of contacting the sour natural gas with a counter current stream of an aqueous DEA solution (usually about 15-30 wt % DEA) at 30-50°C and at 2 SWEET ACID GAS SOUR GAS > ) STEAM Figure 1.1 Flow sheet of the basic amine process 3 pressures ranging from atmospheric to well over 6895 kPa (1000 psi). The acid gases are absorbed by the DEA and the 'rich' DEA is then pumped to an amine stripper. In the stripper the pressure is reduced and the temperature raised to 100-120°C, which is sufficient to remove all but trace quantities of the acid gases. The 'lean' DEA is then recycled back to the absorber. DEA's popularity is based on several factors:- low energy require ments compared with most other solvents, high affinity for acid gases, and resistance to degradation. Degradation is defined as the irrevers ible transformation of DEA into undesirable compounds. Plant experience has shown, however, that degradation does occur and with, at times, sur prising speed. Incidents have been reported in which gas plants have lost their entire DEA inventory within a matter of days. Degradation is undesirable for three main reasons. First, it represents a loss of valuable DEA; second, the degradation products accumulate leading to equipment fouling; third, it is suspected that some degradation products are highly corrosive. Plant operators have tried to solve the problem by changing operating conditions and/or installing activated carbon filters.''" The filters are believed to absorb the degradation compounds. In some cases these measures are successful, but in others they are inadequate for reasons which are not clearly under stood. Furthermore, satisfactory procedures for one gas plant are often ineffective for others. There can therefore be no doubt that the Canadian gas processing industry incurs multimillion dollar costs every year due to DEA losses and increased maintenance resulting from DEA degradation. Consequently, there is a strong incentive to learn how to prevent (or at least minimize) DEA degradation. 1.1 Objectives of the present study Although DEA samples from industrial sweetening units are routinely . analysed to monitor the purity of the solution and to detect degradation compounds, very little has been published in the open literature on DEA degradation. Furthermore the available studies are purely qualitative and no attempt has been made to explain quantitatively the degradation behavior. Thus the main objectives of this study are:-a) To develop a simple analytical technique for analysing degraded DEA solutions. Preferably, the technique should be suited for plant use. b) To isolate and identify the primary degradation compounds. c) To determine the degradation rates as a function of temperature, pressure, and initial DEA concentration. d) To elucidate a reaction mechanism for the production of the various degradation compounds. e) To develop a kinetic model which can predict the degradation of DEA and the production of degradation compounds under typical industrial conditions. f) To propose ways of reducing DEA degradation in industrial sweetening units. g) To investigate ways of purifying degraded DEA solutions and to study the effectiveness of activated carbon. Since DEA degradation results primarily from reaction with C02, the present study emphasizes degradation due to C02 rather than H2S and other compounds such as 02, CS2 and COS. CHAPTER 2 LITERATURE REVIEW The DEA sweetening process has been widely accepted for removing 2-7 C02 and H2S from natural gas. There have been several studies mainly 8—11 directed at improving plant performance and removing degradation 112 8 products. ' ' In addition there are several handbooks available which review natural gas processing and present general analytical methods for 13-17 routine analysis of gas treating solutions. However, there are surprisingly few references dealing specifically with DEA degradation and the analysis of the degradation products. In fact most of the experimental studies have been limited to the absorption of C02 into DEA which takes place in a matter of seconds whereas degrada-• j * *u 18-29 tion takes place over a period of months. 2.1 Absorption of C02 in aqueous DEA solutions The mechanism of C02 absorption, unlike that of H2S, is not simple and involves the establishment of several equilibrium reactions. Although the literature on C02 absorption is extensive, the absorption is still not fully understood. When a C02 molecule dissolves in water, due to its acidic chemical characteristics it first hydrates to form carbonic acid, H2C03 (see Eq. 2.1. in Table 2.1). The H2C03 molecule, in turn, slightly dissociates to form hydrogen (H+) and bicarbonate (HC03 ) ions. The bicarbonate ion 6 Table 2.1 Equations governing C02 absorption in aqueous DEA solution (R denotes C2Hit0H) Acid-base reaction C02 + H20 ^ H2C03 [2.1] H2C03 ^ HC03~ + H+ [2.2HC03~ ^ C03~~ + H+ [2.3] R2NH + H20 ^ZT R2NH2+ + 0H~ [2.4R2NH + H+ ^± R2NH2+ [2.5] C02 + OH" ^± HC03~ [2.6R2NH2+ + HC03~ ^ [R2NH2+] [HC03~] [2.7] 2R2NH2+ + C03" ^ [R2NH2 + ]2 [C03~] [2.8[R2NH2+]2 [C03~~] + C02 + H20 2[R2NH2 + ] [HC03~] [2.9] Carbamate formation R2NH + C02 R2NH+COO~ [2.10] R2NH+C00~ + H20 R2NC00" + H*0 [2.11R2NH+C00~ + 0H~ ^ R2NC00~ + H20 [2.12] R2NH+C00~ + R2NH R2NC00~ + R2NH2 + [2.13R2NC00~ + H+ + H20 ^ R2NH2+ + HC03~ [2.14] 7 and the DEA molecule may then react thereby forming the DEA bicarbonate (see Eq. 2.7). The solubility of C02 in water is thus enhanced by the presence of DEA. This is due to DEA being alkaline and forming hydroxyl ions (OH ) with water. The OH can then also react directly with C02 to form HC03 (see Eq. 2.6). The overall rate of C02 absorption by these acid-base reactions is quite slow, especially when compared to the absorption of H2S. The difference between these rates of absorption has led to the development of amine processes that are able to selectively 30,31,32 absorb H2S. A second set of C02 absorption reactions also occurs involving the labile hydrogen atom present in the DEA molecule. In this case the C02 molecule reacts directly according to Eq. 2.10 with the DEA mole cule to form a zwitteron complex, i.e., a dipole molecule which is extremely 23 unstable. The zwitteron is then rapidly deprotonated by water, OH or another amine molecule (see Eqns. 2.11 to 2.13), to form the DEA carba mate, R2NC00 . The rate of C02 absorption via the carbamate reactions is much faster than the C02 acid-base reaction, but still somewhat slower than the H2S absorption reaction. 21 . . Astarita et al. proposed that the enhancement of C02 solubility in water by DEA is due to a rather complex "mass transfer with chemical reaction" mechanism. In the region near the gas-liquid interface, C02 was postulated to dissolve in the water and undergo the fast carbamate reactions. The reactions were believed to be fast enough to cause a steep C02 concentration profile near the gas-liquid interface and there by increasing the mass transfer rate. The carbamate ion diffuses into the bulk of the liquid where it reverts to bicarbonate and liberates free amine (see Eq. 2.14). The amine is then able to diffuse back to 8 the interface region where it can react with additional C02 in an increased solubility of C02 by converting C02 to HC03 tative behaviour is shown in Fig. 2.1. This results The quali-Gas phase Gas-1iquid Interface Bulk of liquid phase CO, CO, Fast reaction with amine to carbamate Slower reaction C02 and hydroxyl ion Amine diffusion Carbamate diffusion Slow reversion of carbamate to amine and bicarbonate HCO, Fig. 2.1 Shuttle mechanism for C02 absorption into aqueous amine solutions 33 Jorgensen has proposed an additional reaction between C02 and DEA. In this case the C02 reacts with an alcohol group in the DEA mole cule to form an alkyl carbonate. RNH-C-HijOH + OH RNH-C2H40 + H,0 [2.15] RNH-C.H^O + CO. RNH-C,H„-0-C \ [2.16] However, this reaction occurs only in strongly alkaline solutions (pH 26 > 13) and at low temperatures. It is therefore unlikely to be of importance in natural gas treating units. Several authors have also + 24 25 29 proposed the existence of the R2NCOO R2NH2 complex. ' ' However, in aqueous solutions this complex can be considered to be almost com pletely dissassociated. Thus the gas treating solution can be considered to be a complex mixture of ionized species in equilibria consisting mainly of H+, OH , HC03~, R2NC00~' R2NH2+, as well as the molecules C02 and R2NH. 2.2 DEA degradation Besides the establishment of the ionic equilibria within the gas treating solution, there are certain side reactions by which DEA and - + * DEA carbamate (R2NC00 H ) undergo further change. DEA, in general, is not easily recovered from these compounds and these side reactions are termed the "degradation reactions." DEA degradation is a complex phenomenon. Smith and Younger ' ^ ' ^ as well as Nonhebel^ have reported that degradation apparently depends on temperature, pressure, gas composition, amine concentration, pH, and the presence of metal ions. However, the quantitative relationships between these variables and degradation have not been reported. It is, therefore, impossible to predict DEA degradation rates or, alternatively, to estimate improvements from changes in operating variables. The situa tion is further complicated by the fact that the degradation products are large organic molecules which are difficult to detect and identify. * — + DEA carbamate is written in the ionic form, R2NC00 H , since in aqueous solutions it can only exist as ions. Carbamic acid, R2NCOOH, is extremely unstable and reverts to DEA and C02; R2NC00H has never been isolated. 10 34 Polderman and Steele were the first to publish a comprehensive investigation on DEA degradation in 1956. . Their work consisted essen tially of placing a 25 wt % DEA solution into a pressure vessel, saturat ing it with C02 at 25°C, sealing the vessel and raising the temperature to between 100-175°C. The pressure inside the vessel ranged from 1257.3-4137kPa (180-600 psi). After eight hours the vessel was cooled to room temperature and the contents analysed by fractional distillation and crystallization. The DEA loss ranged from 0% at 100°C and 1257.3 kPa (180 psi) to 97% at 175°C and 4137 kPa (600 psi). They discovered N,N-bis(2-hydroxyethy1)piperazine, or BHEP, in the degraded solutions and suggested the following reaction for its formation. 0 II H0-C,Hu C \ / \ N-H + C02 —*• H0-C2H4 - N 0 [2.17] + H20 / 2 2 I I H0-C2Hu CH2 CH2 DEA HEOD 0 II C v CH, - CH. / \ / \ 2 HO-C.Hi, - N 0 —• H0-C,H„ - N N-C,H„-0H I I \ / CH2 CH2 CH2 - CH2 HEOD BHEP + 2C02 [2.18] HEOD or 3-(2-hydroxyethyl)-2-oxazolidone was regarded as an intermediate and somewhat unstable compound. The authors also noted the presence of other degradation products but did not characterize them owing to the lack of suitable analytical techniques. Equations 2.17 and 2.18 indicate that C02 acts like a catalyst, i.e., it is neither consumed nor formed. If the reaction scheme is correct, then DEA degradation is governed by a first order kinetic equa tion provided C02 is present in excess. Since DEA and C02 are primarily present as ions in aqueous solutions, it is unlikely that the above reac tions are realistic. 35 Using more sophisticated analytical techniques, Hakka et al. were also able to detect N,N,N'-tris(2-hydroxyethyl)ethylenediamine, or THEED in degraded DEA solutions. However, they did not propose a reaction mechanism for its formation. According to Hakka et al., THEED occurred frequently in gas treating solutions and was one of the major degradation compounds. 2 3 36 These authors and others ' ' found that both BHEP and THEED have acid gas removal properties virtually identical to those of DEA. How ever, under gas treating conditions, only one of the nitrogen atoms in the BHEP or THEED molecule is likely to combine with the acid gas. Hence, on a weight basis, the capacity of the treating solution falls with increasing DEA degradation. 12 41 Smith and Younger and others have discussed DEA degradation and mentioned several other degradation compounds. One of these com pounds was found to have the same retention time as triethanolamine, or TEA, in the gas chromatographic analysis. In many cases the DEA treating solution contains small amounts of monoethanolamine, or MEA. This compound can also degrade38,39'^0 forming oxazolidone (OZD), l-(2-hydroxyethyl)imidazolidone (HEI), N,N'-bis(hydroxyethyl) urea (BHEU), and N-(hydroxyethyl)ethylenediamine (HEED). 36 In a recent study by Blanc et al. the authors reacted DEA with C02 and in another experiment with HEOD. Both experiments were conducted in a sealed autoclave at temperatures of 90-130°C. They proposed various reaction mechanisms for the production of HEOD, THEED, and BHEP and other degradation compounds. However, they provided no quantitative data in support of these mechanisms. 41 42 Choy ' performed a se_ries of controlled experiments and found that DEA degradation appears to be first order between 165-185°C and at C02 pressures ranging from 1207-4137 kPa (175-600 psi). Degradation rates were, however, affected by initial DEA concentration, which cannot be explained in terms of a simple first order mechanism. Furthermore, several degradation compounds were detected, and although their chemical structure was not determined, their concentration changes with time sug gested a series of simultaneous and consecutive degradation reactions. 43-45 Recent work by Kennard and Meisen have confirmed that DEA degradation is not a simple first order reaction. There is another type of degradation compound which are called heat stable salts. These salts occur when a stronger acid than H2S or C02 reacts with the amine forming an unregenerable salt, i.e., DEA is not easily recovered by the action of increasing the temperature since the bond is too stable. These stronger acids were described in the 46 47 36 early work by Henry and Grennert. ' Blanc et al. later identified them as formic, acetic, propionic, and oxalic acids. The production of these acids has been attributed to the presence of oxygen although the mechanism for production is not clearly understood. Since the present study is concerned with the reaction between DEA and C02, heat stable salts will not be considered further. 7 34 Degradation compounds of high molecular weight have been proposed ' but not identified. These compounds may be linear polycarbamides con taining polyalkylene amine structures. Other studies have been mainly concerned with the effect of degrada tion products on corrosion. ^' ' ' DEA itself is not considered corrosive, but degraded DEA solutions can attack mild steel. It has 36 been suggested, however, that since the pH of 30 wt % to DEA is in the range of 11.5-10 at temperatures of 20-100°C then the corrosion of mild steel becomes impossible. Further, it has been shown that the 2 major degradation compounds BHEP and THEED are relatively noncorrosive. ' 3 34 35 36 ' ' ' The corrosion may, therefore, be caused by other trace impurities such as cyanides, chlorides, or the organic acids. 2.3 Analysis of DEA and its degradation products A quantitative study on DEA degradation is dependent on a reliable analytical procedure for measuring the degradation compounds. The anal ysis of DEA and its degradation compounds has proven to be rather diffi cult because the degradation products tend to: a) have fairly low vapour pressures; b) decompose at elevated temperatures; c) be highly polar; and d) occur in low concentrations. There is another problem which arises from the fact that degraded solu tions are a complex mixture of ionized species in equilibrium. The 14 process of analysis may have the effect of shifting the equilibrium and it therefore becomes impossible to isolate and measure each component in the form in which it exists in the degraded solution. For example, it is impossible to isolate the carbamate R2NCOOH since it readily decom poses to C02 and DEA. 46 44 Henry and Grennert ' were amongst the first researchers interested in the detection and measurement of DEA salts in refinery samples. They investigated four types of acidic materials: a) organic acids, b) chlor ides, c) cyanides and thiocyanates, and d) sulphites, sulphates, and thiosulphates. They used mainly potentiometric titration for the DEA analysis. They also discussed conventional wet chemical methods such as titration and kjeldahl total nitrogen determination, as well as other methods for the determination of total sulphur, sulphide, mercaptide, sulphate, thiocyanate, cyanide, chloride, carbonate, alkalinity, and sodium. This study was however limited because it failed to detect organic degradation compounds. The "Gas Conditioning Fact Book," pub lished by Dow Chemical Company^ provides a description of conventional wet chemical methods for testing DEA samples. However, these methods are again unsuitable for identifying DEA degradation compounds. A comprehensive study on the analysis of DEA gas treating solutions 37 was produced by Gough. Here an attempt was made to describe anal ytical schemes that would lead to a useful interpretation of quality. Two schemes were described: a) a comprehensive scheme for component analysis, used when detailed information on composition is required; b) a simple scheme for quality evaluation, useful on a routine basis and providing information required for routine plant operation. Unfortunately this study was also not suitable for observing and identifying 15 the individual degradation compounds. 48 Brydia and Persmger described a chromatographic technique for analysing ethanolamines. Because direct chromatographic procedures were not entirely successful (excessive peak tailing due to strong hydro gen bonding), derivatization prior to chromatographic separation was investigated. Trifluoroacetyl anhydride was used to convert non volatile amines into volatile amine trifluoroacetyl derivatives. The authors experienced problems with reproducibility, precision, and the presence 49 . . of water. Piekos et al. eliminated these shortcomings by converting the alkanolamines to trimethylsilyl derivatives. N,0-bis(trimethylsilyl) acetamide was used which reacts with both the amino and hydroxyl groups of the alkanolamines. The process is called silylation and produces fairly stable compounds which are more easily separated and identified by gas chromatography. The addition of a trimethylsilyl group also decreases the polarity of the alkanolamine and reduces hydrogen bonding. Silylated compounds are more volatile and more stable due to the reduction of reactive sites. Successful separation of MEA, DEA, and TEA derivatives was conducted and the presence of up to 5% of.water could be tolerated, provided there was a large excess of silylating agent. A recent paper by Saha et al.^° investigated problems arising from converting the amines to stable derivatives prior to analysis by gas chromatography. For example, derivative preparation is time consuming, the derivative reactions may be incomplete and the derivatives may not be stable for long periods. Consequently the use of organic polymer beads as the column packing for G.C. analysis of alkanolamines was inves tigated. The authors found that Tenax G.C.,"^ which is a porous poly mer based on 2,6-diphenyl-paraphenylene oxide, was able to separate alkanolamines with excellent results. The authors were able to separate an aqueous mixture of MEA, DEA, and TEA in less than eight minutes using a 1/8" O.D, 4' long stainless steel column. No sample preparation was required and the column was unaffected by the presence of water. Probably the only study dealing specifically with the analysis of 52 DEA and its degradation products was that performed by Choy and Meisen. Their technique consisted of first drying the degraded DEA sample by air stripping, dissolving it in dimethyl formamide and silylating it with N,0-bis(trimethylsilyl)acetamide. The silylated compounds were then separated using a 1/8" O.D, 6' long stainless steel column packed with 8% 0V17 on 80/100 mesh chromosorb followed by flame ionization detec tion. Although the method was reliable and accurate, it was time con suming and unsuitable for plant use. In particular, the silylation stage required considerable care particularly with regard to the removal of water. Other methods for the analysis of amines and amine related compounds 53 54 have been reported. These studies include paper chromatography, ' salting out chromatography,and thin layer chromatography ."^ All these methods suffered from excessive tailing and lack of reproducibility. Also none of these methods have been applied to DEA degradation compounds. CHAPTER 3 DEVELOPMENT OF THE ANALYTICAL TECHNIQUE . Before a study of DEA degradation could be undertaken a reliable, quantitative method of analysis of DEA and its degradation products had to be developed. The method should be rapid, highly sensitive, and require minimal sample size and preparation. Furthermore it is desirable for the method to be applicable for industrial as well as laboratory use. Many methods have been investigated for the analysis of DEA and its degradation products such as wet chemistry, infra red and ultra violet spectroscopy, paper and thin layer chromatography. However, they all tend to have drawbacks such as being generally inaccurate, non specific, unreliable, and expensive. In addition, these methods tend 41 to be slow and inconvenient. Choy stated that gas chromatography was probably the most promising analytical technique. As mentioned before, one of the problems of analysing DEA and its products is their low vapour pressure. This requires the use of high injection and column temperatures. However poor thermal stability of DEA leads to problems with the reproducibility of measurements. Also, the polar hydroxyl and amino groups have a strong affinity for most column packings. This results in long elution times, large peak broadening, and peak assymetry. The presence of water in the sample also creates problems since only a few column packings can tolerate aqueous samples. 18 3.1 Gas chromatographic technique 52 Choy's method using derivative gas chromatography although reliable, was felt to be too time consuming and unsuitable for plant use. This was due to the complicated sample preparation since the silylation reactions were extremely sensitive to water. Furthermore, there is the problem of incomplete silylation of all compounds. Silylation of the hydrogen 54 bound to the nitrogen atom of alkanolamines is known to be difficult. An attempt was therefore made to find a simpler and more direct technique for analysing DEA and its degradation compounds. A thorough review of the literature yielded an article by Saha et al.^ who used Tenax G.C. to separate alkanolamines. Tenax G.C. is a porous polymer based on 2,6-diphenyl-paraphenylene oxide which has a weakly interacting surface and can be used at temperatures up to 450°C. According to the manufacturers, columns may be operated for several weeks at temperatures up to 400°C without significant baseline drift and decomposition of the packing. Since the organic polymer beads are solids, mass transfer is rapid and fast elution and sharp peaks are obtainable. Using a temperature programmable gas chromatograph (Hewlett Packard Model 5830A), a 1/8" O.D, 6' long stainless steel column packed with Tenax G.C. (purchased from Alltech Associates, Illinois) was tested and found to be successful. Two other packings, which could tolerate aqueous samples, were also tested; these being 4% Carbowax 20m on 60/80 58 mesh Carbopack B (purchased from Supelco Inc., Penna.) and 28% Pennwalt 59 223 + 4% kOH on 80/100 mesh Gas-chrom R (purchased from Applied Science Lab. Inc., Penna.). Both columns were found to be unsuitable due to either low sensitivity on excessive peak tailing. 19 3.1.1 Evaluation of the Tenax G.C. column. Initial tests were performed with solutions made by mixing distilled water with reagent grade MEA, DEA, and TEA. Using a flame ionization detector, nitrogen as the carrier gas and temperature programming, excellent separation was obtained. From the literature review it was apparent that the main degradation compounds are HEOD, THEED, and BHEP. Unfortunately only BHEP could be obtained commercially. Therefore standards for HEOD and THEED were prepared in the laboratory and their synthesis is described in chapter 4. All three compounds were easily separated with the Tenax column. A paper by Alltech Associates Inc.^ indicated that stainless steel causes ethanolamine to undergo catalytic degradation. However, no evidence of degradation within the column was observed for any of the compounds tested. Although the inlet port and column were operated at high temperatures (up to 300°C), there was no observable thermal decom position of the tested compounds. This was confirmed by the sharp, sym metric, and highly reproducible peaks. 3 ."1.2 Operating conditions. After several initial trials, optimum conditions were found for the separation of DEA and its degradation pro ducts. Table 3.1 summarizes the facilities and operating conditions used for the separation. Temperature programming was used in order to achieve a good separation of all degradation compounds, since these compounds varied considerably in molecular weight and polarity. The maximum temperature of 300°C was adopted to ensure that all compounds were volatilized. However, even at this temperature it is possible that some of the very heavy degrada tion compounds, such as the polylinear carbamides did not elute. Usually a luL sample in conjunction with an attentuation value of 13-14, was found Table 3.1 Analytical equipment and operating conditions for G.C. analysis Gas Chromatograph Manufacturer Model Detector Chromatographic Column Material Dimensions Packing Operating Conditions Carrier gas Injection port temp. Detector port temp. Column temp. Syringe Manufacturer Model Injected sample size Hewlett Packard 5830A H2 flame ionization Stainless steel 1/8" O.D, 6' long Tenax G.C., 60/80 mesh N2 at 25 ml/min 300°C 300°C Isothermal at 150°C for 0.5 min., then temperature raised at 8°C/ min to 300°C Hamilton Co. 701, lOpL, with fixed needle and Chaney adapter luL 21 to be suitable for the detection and separation of all compounds. How ever, in some cases the sample size was increased to 2uL or more to detect-trace quantities of degradation compounds. 3.2 Analytical procedure and performance Typically l.OpL samples of the degraded DEA solution were injected directly into the column with a precision syringe fitted with a Chaney adapter. The adapter was used to ensure that a constant volume of sample was injected into the column. To improve the accuracy, a needle guide was used at the injection port. This guide not only protects the fragile syringe needle, but serves as a spacer for needle penetration and lengthens the septum life by using a single hole for repeated injections. Needle guides were found to be indispensible for high precision work. The analysis was usually performed for a period of 30 minutes in order to ensure the elution of heavy compounds. After each run the column had to be cooled from 300°C to 150°C which took about 5 minutes. Therefore, a complete analysis required a total of about 35 minutes. DEA and known degradation products could be detected accurately down to concentrations of about 0.5 wt %. The reproducibility was excel lent (typically within ± 5.0% with a new column) and peak tailing and baseline drift did not represent special difficulties. Figures 3.1 and 3.2 show examples of the separation achieved with the Tenax G.C. column. 3.2.1 Column performance. The column itself required no special care and was conditioned simply by passing nitrogen through it at its maximum operating temperature (300°C) for 8-10 hours. The column has a fairly long life; for example, one column was in continual use for nearly a year. However, when a column fails, it fails rapidly and 22 MEA DEA TEA Figure 3.1 Typical chromatogram for MEA, DEA, and TEA Figure 3.2 Typical chromatogram of a degraded DEA solution 23 becomes incapable of separating the heavy compounds. It is likely that the column becomes clogged with the polylinear carbamide compounds, which • probably never leave the column. Thus the more degraded the sample the shorter the column life. In some cases a 9' column was used instead of the standard 6' column. This improved the separation of degraded compounds, especially the separ ation of BHEP and HEOD. Also it allowed a direct comparison to be made 36 between the results of this study and those of Blanc et al. who used a 9' Tenax G.C. column. However, the longer column increased the elution times, and most analyses were therefore performed using the 6' column. 3.3 G.C. calibration Calibration plots of concentration versus peak area were produced simply by injecting known concentrations of the various degradation com pounds into the chromatograph and noting the peak area which was auto matically calculated by the chromatograph's computer. At least five injections were made for each concentration and the peak area averaged. Figures 3.3 to 3.6 show the calibration for DEA, HEOD, BHEP, and THEED. This form of calibration did not use an internal standard and is termed 'direct calibration'. 3.4 Maintenance of chromatographic equipment Generally very little maintenance is required. Basically the chromatograph and syringes must be kept clean. In some cases deposits tend to build up in the injection port and have to be removed. Further more, deposits accumulate on the detector jets and can result in excessive spiking (or noise) on the chromatogram. The cleaning of the flame ioniza tion detector is difficult and the removal of the probes is not recommended unless absolutely necessary. An easier method is to inject, 10-30uL Figure 3.4 Calibration plot for HEOD BHEP PEAK AREA Figure 3.5 Calibration plot for BHEP Figure 3.6 Calibration plot for THEED 28 of Freon 113 flame ionization detector cleaner (purchased from Supelco, Inc.) with equipment operating under normal conditions. Freon elutes from the column and produces hydrogen fluoride as the cleaning agent when burnt in the hydrogen flame. Since the column does eventually wear out it is considered good practice to check the calibration at least once a month with standard samples. If there is considerable disagreement between the calibration curve and the analysis of the standard samples, then the column should be replaced. Septa should be replaced at least every 50 injections since they tend to accumulate deposits and eventually begin to leak. No other routine maintenance is required except keeping the machine clean and free of dust. The grill to the fan which cools the circuit boards may get clogged with dust and restrict the flow of cooling air causing overheating of the circuit boards. Thus the grill must be checked periodically. The printer is generally trouble free requiring only cleaning of the slide rod for the printer head and keeping it free of oil and grease. 3.5 Advantages of the analytical technique The advantages of the present analytical technique can be summarized as follows. 1. No sample preparation required. 2. Water has no effect on the column. 3. Very simple. 4. Long column life. 5. Small sample required. 6. Reliable and reproducible. 7. Speed, i.e., analysis is completed in less than 35 minutes. 8. Suitability for plant use. 3.6 Errors The major source of error arises during sample injection. Since direct calibration was used to determine the concentration, then each sample injection had to be as identical as possible. For example, increas ing the injection time usually resulted in slightly larger peak areas, since a small volume of liquid, otherwise held in the needle, is partially vapourized and enters the column. Another problem with direct calibration is the sensitivity of the detector is assumed to remain constant from day to day. This assumption is only valid provided the detector is kept clean. Another source of error is changes in the flow of carrier gas. As the column becomes clogged, the flow tends to fall unless adjusted. Therefore, it is best to check the flow of carrier gas daily. The only other noticeable error is concerned with the calculation of the peak areas which the chromatograph performs automatically. As long as the peaks are sharp and symmetrical and there is little base line drift, there is usually no problem. If the peaks tend to tail or bunch, the automatic integrator may make small errors in deciding where to begin and end integration. In general, this form of error is minor compared to that produced by sample injection. 3.6.1 Accuracy. The accuracy of the technique can be simply cal culated using the relative standard deviation oD. 30 where N = number of measurements x = measured value x = arithmetic mean a = standard deviation For example, for the analysis of a sample of DEA the following six con centrations were recorded for a sample of aqueous DEA of known concentra--3 tion 3.5 x 10 moles/cc.:-3.54 x 10~3, 3.45 x 10"3, 3.52 x 10~3, 3.51 x 10~3, 3.55 x 10_3, 3.43 x io"3 x = 3.5 x 10~3 o = 2.19 x 10~5 oR = 6.3 x io"3 Similarly the following five concentrations were recorded for a sample -3 of aqueous DEA of known concentration 0.95 x 10 moles/cc.:-9.25 x io"4, 9.75 x io"4, 1.01 x 10~3, 9.81 x 10~4, 9.35 x 10_4 x = 9.652 x 10~4 o = 1.745 x 10~5 oR = 1.81 x 10~2 There is of course one final source of error and that occurs when reading the calibration curves. Therefore considering all the error sources and the relative standard deviation, the accuracy of the analytical technique has been found to be ± 5%. 3.7 Units of DEA concentration DEA concentration is, throughout this study, expressed in units of wt % or moles/cc. It should be noted that since the molecular weight of DEA is 105, the concentration expressed as moles/cc can roughly be _2 converted to wt % by multiplying by 10 ; e.g., 1.5 * 10 moles/cc is roughly equivalent to 15 wt %. CHAPTER 4 SYNTHESIS OF SELECT DEGRADATION COMPOUNDS FOR CALIBRATION OF THE GAS CHROMATOGRAPH In order to study DEA degradation quantitatively it is necessary to measure the concentration of DEA and its major degradation products in solution. In order to calibrate the chromatograph, standards of the various degradation compounds had to be obtained. From previous studies the major degradation compounds were thought to be HEOD, THEED, and BHEP. Unfortunately, only BHEP was available commercially. HEOD could only be obtained from ICN Pharmaceuticals, Inc. but its purity was found to be too low for calibration. It is likely that HEOD reverts slowly to DEA and this is discussed further in chapter 11. THEED was unavailable from any commercial source. It was, therefore, decided to synthesize both HEOD and THEED in the laboratory. 4.1 Synthesis of HEOD A thorough search of the literature, back to 1920, revealed only the following methods for synthesizing 2-oxazolidones: a) From alkanolamines and phosgene.^ The alkanolamine is reacted with phosgene in chloroform in the presence of lead carbonate which neutralizes the hydrochloric acid produced in the reaction. 31 0 II RNH - C.Hi, - OH + COCL, • R-N 0 + 2HCL [4.1] I I CH2 CH2 62 63 b) From alkanolamines and dialkyl carbonate. ' 0 II R'O C \ NaOH / \ RNH - C2H4 - OH + C = 0 • R-N 0 [4.2] / -2R'0H * | | R'O CH2 CH2 NaOH acts as a basic catalyst. 64 c) From 8 halogenalkyl carbamate. The carbamate is boiled in an aqueous KOH solution. 0 II 0 C II KOH / \ RNH - C - 0 - C-H^-CL • R-N 0 + KCL + H,0 [4.3] I I CH2 CH2 d) From ethylene oxide (EO) and a hydroxyl alkyl isocyanateThe oxide is heated with the isocyanate in the presence of potassium iodide or lithium chloride, which act as catalysts. 0 C / \ / \ CH2 - CH2 + R-N=C=0 • R-N 0 [4.4] I I CH2 CH2 e) From ethylene oxide (EO) and 2-oxazolidone (OZD).^^ Equal amounts of ethylene oxide and oxazolidone are heated in the presence of a trace amount of water. 33 0 0 0 C C / \ / \ / \ CH2 - CH2 + H-N 0 —«- H0-C2H^ - N 0 [A.5] 11 II CH2 CH2 CH2 CH2 With the exception of (e) the methods were non-specific for the production of HEOD. Furthermore complete details of the reaction condi tions were not stated. Methods (b) and (e) were attempted with partial success. Using method (b) the following synthesis was performed. Equal amounts of ethyl carbonate and diethanolamine were mixed and 15g charged to a 25ml. pressure reactor together with about 5g of IN sodium hydroxide. The reactor was sealed and pressured to about 689.5 kPa (100 psi) with nitro gen. The reactor was then placed in a water bath and heated to about 50°C for one hour. The results of this experiment were rather disappoint ing. HEOD was produced only in low amounts since there were many side reactions taking place. It was felt that purification would prove diffi cult as well as time consuming and therefore method (e) was tried. Equal amounts of ethylene oxide and oxazolidone were mixed and 15g charged to the 25 ml. pressure reactor before adding about lg of water. The reactor was sealed and pressurized to about 689.5 kPa (100 psi) with nitrogen. The reactor was placed in a water bath and heated to 70°C for about eight hours. HEOD conversions of about 70% were achieved with MEA and DEA being the main impurities. However, further purifica tion proved once again to be very difficult. Several methods were tried such as distillation; solvent extraction with chloroform, acetonitrile and pyridine; and gel chromatography. None of these approaches were 34 very successful. Since HEOD was required in a purity of at least 95%, for calibration of the chromatograph, it was decided to enlist the help of a small firm specializing in custom synthesis of rare chemicals (Syn-thecan Lab., Inc., Vancouver). A pure 50g sample was purchased and used for the calibration. 4.2 THEED synthesis Information with regards to synthesis of THEED proved very diffi cult to find and a thorough literature survey revealed only one relevant reference.^ Three different methods were tried for the synthesis of THEED and are summarized below. a) From ethylenediamine (ED) and chloroethanol. The reaction was expected to produce the following compounds:-H2N-C2H^-NH2 + HOC2H^CL • H0C2 Hu-NH-C2 Hu-NH2 (HEED) + (HOC2Hu)2-N-C2Hu-NH2 (BHEED) + (HOC2Hu)2-N-C2Hu-NH-C2HuOH (THEED) + (HOC2H4)2-N-C2Hu- N-(C2HuOH)2 (TEHEED) + HCL [4.6] The reaction was initially performed by mixing ethylenediamine with chloroethanol, in the molar ratio of 1:3, in a glass beaker and heating the mixture to about 100°C. Unfortunately the reaction was found to be highly exothermic and a very viscous yellow substance was produced. Analysis showed that no THEED was produced. A similar experiment con ducted at 50°C resulted in no reaction taking place. A more controlled experiment was performed where 15 cc. of the reaction mixture was placed in the 25 ml. pressure reactor and the reactor pressurized to about 689.5 kPa (100 psi) with nitrogen. The reactor 35 was then heated to 90°C in a water bath for about 1 hr. The reaction produced numerous compounds one of which may have been THEED. Purifica tion would have been too difficult and therefore other methods were tried. b) From ethylenediamine and ethylene oxide. This reaction was far less exothermic and easier to control. The expected products were similar to those stated in Eq. 4.6. 15g of a mixture containing 3 moles of ethylene oxide to one mole of ethylenediamine were charged to the 25 ml. pressure reactor. About 68 lg of water was added as a catalyst. The reactor was sealed and pres sured to 689.5 kPa (100 psi) with nitrogen and heated to about 50°C in a water bath for one hour. Five distinct compounds were produced (see Fig. 4.1). Two of the peaks corresponded to the two isomers of BHEED, namely (H0C2 Uk) 2 -N-C2 H^-NH., and (HOC2Hu)-NH-C2Hl4-NH-(C2H1<OH). The substituted ethylene diamines were produced in the following amounts: HEED 17.7%, BHEED 32.4%, THEED 35.8%, and TEHEED 14.1%. Since all these compounds had similar characteristics it was felt once again that purifi cation would present problems especially since THEED was present in only 36% purity. Thus a third method was tried. c) From diethanolamine and N-(2-hydroxyethyl)ethylenimine (HEM). The following reaction was expected to occur:-CH, (H0C2H.J2NH + H0-C2H«, •= N | • (HOC2H«t),-N-CaHi,-NH-C2HitOH [4.7] CH2 Approximately 200 cc. of an equimolar solution of DEA and HEM was charged to a 600 ml. stirred autoclave (details of the autoclave are given in chapter 6). The 600 ml. autoclave was used so the reaction could be followed more closely by removing samples while the reaction 36 Figure 4.1 Chromatogram of substituted ethylenediamines was still being carried out. About 5g of aluminium chloride were added to the reaction mixture. The autoclave was sealed and pressurized to 689.5 kPa (100 psi) with nitrogen and heated to 120°C for about eight hours. THEED was produced at about 78% purity, the major impurity being DEA. However, even this concentration was not sufficiently high for calibrating the gas chromatograph. After trying several methods of purification it was found that gel chromatography was most suitable and it was possible to produce THEED of 95% purity. A 2 cm. diameter, 40 cm. long glass column was used for the gel chromatography. The column was packed with 60-200 mesh silica gel. Since THEED and DEA are both soluble in water, water was used as the solvent. A trace of ammonium hydroxide was found to aid separation. lOcc. samples of impure THEED were charged to the column and an elution rate of about lcc. per 10 minutes was established. Fractions were collected and analysed for THEED content. When sufficient THEED had been collected the samples were concentrated by boiling off the water thereby leaving a viscous colourless liquid of THEED. CHAPTER 5 IDENTIFICATION OF DEGRADATION COMPOUNDS 5.1 Identification using the gas chromatograph Before proceeding to the study of degradation reactions the compounds responsible for the peaks on chromatograms had to be identified. For a typical run (i.e., run number 3, where 30 wt % DEA was degraded for eight hours under A137 kPa (600 psi) C02 at 205°C) over 20 peaks were observed. To identify these peaks where possible the following method was used. If a compound was suspected of being produced (based on infor mation from the literature or otherwise), then it was either purchased or synthesized in the laboratory (see chapter A). A known concentration of the compound was injected into the chromatograph and its retention time noted. This retention time could then be compared with the elution times of the degraded sample of DEA. If a degradation compound produced a peak with the same retention time as that of the standard compound it could be inferred that the degradation compound and standard compound were the same. However, this method does not give a completely reliable identification of a compound since there can be more than one compound with the same retention time. For example, a peak occurred after about 12 minutes which is also the retention time for TEA. It is unlikely, however, that TEA is a degradation product and it is probable that the peak is caused by another compound. TEA does exist as an impurity of 38 39 1-2 wt % in the DEA solution. However, the peak area is usually much greater than that produced by TEA in the concentration range of 1-2 wt %. Thus further information is required for the positive identification of a degradation compound. It was found that using a gas chromatograph and a mass spectrometer (GC/MS) was best suited for this task. 5.2 Identification using a GC/MS The mass spectrometer simply vapourizes a compound and produces ions from the neutral molecules by bombarding the vapour with electrons. The ions are formed into a beam and accelerated through the field of a powerful electromagnet. The ions are forced into a circular path and become separated according to their mass to charge ratio. The different ions are detected by an electrometer which measures the charge collected on current carried by the ions. Recording the changes in charge a spectro graph is produced of ion current versus mass number. Thus each peak on the spectrograph corresponds to a different ion. Since the molecule fragments in the same manner under similar conditions, the mass spectro graph provides a characteristic "thumbprint" for each compound. The use of a gas chromatograph with the mass spectrometer makes the GC/MS a very useful tool. Samples of the degraded DEA solution were injected into the GC which separated the compounds. Then each compound flowed into the MS and produced a mass spectrograph corresponding to each peak on the chromatograph. A Hewlett Packard GC/MS (Model 5985B) was used to produce mass spectrographs of all suspected compounds. Only the mass spectrographs of MEA, DEA, and TEA could be found in the available registries of mass spectral data. Therefore, standard mass spectrographs were made from AO pure samples of DEA degradation compounds. These standard mass spectro graphs are given in Appendix F. Samples of degraded DEA solutions could then be analysed using the GC/MS. The resulting spectrographs were compared with the standard spectrographs enabling a positive identification of a degradation compound. For example it was suspected the compound producing the peak marked by the arrow in Figure 5.1 was HEOD. The mass spectrographs for HEOD and the unknown compound are shown in Figs. 5.2 and 5.3. As can be seen they are very similar. The most notable similarity being the two peaks of masses 100 and 101, which are characteristic of HEOD. Therefore HEOD could be positively identified. 5. 3 Identified degradation compounds Using the methods previously described it was possible to identify 14 compounds in degraded DEA solutions (e.g., Fig. 5.4 shows a typical chromatograph of a degraded DEA solution). Many other compounds were detected, but since their concentrations were extremely low their identi fication was considered not worth pursuing. Table 5.1 is a summary of the compounds detected with their reten tion times using the Tenax G.C. column and conditions described in chapter 3. Mass spectrographs for these compounds are found in Appendix F. Possible mechanisms for their production will be discussed in chapter 11. Figure 5.1 Chromatogram showing peak of unknown compound ice so i 41 4 0 ioo 1 I, l.il r ieo 131 1S4 ' 1 120 —1 • 1 140 Figure 5.2 Mass spectrum of HEOD Figure 5.3 Mass spectrum of unknown compound DEA Figure 5.4 Typic al chromatogram of a degraded DEA solution Table 5.1 Compounds found in degraded DEA solutions Compound Retention time min.* MEA 1.4-1.5 HEM 3.1-3.5 HEED 6.8-7.2 DEA 7.4-8.0 HEP 9.8-10.2 OZD 10.4-10.6 TEA 12.0-12.5 BHEED 12.8-13.2 BHEP 13.0-13.6 HEOD 13.4-14.0 HE I 15.5-16.0 THEED 17.2-17.4 BHEI 18.2-18.5 TEHEED 20.4-20.6 *This will vary slightly according to the concentra tion of the compound present in the sample and the age of the column. CHAPTER 6 EXPERIMENTAL EQUIPMENT AND PROCEDURE FOR THE CONTROLLED DEGRADATION OF DEA 6.1 600 ml.autoclave Since DEA degradation is rather complex, the experiments had to be performed under carefully controlled conditions. In particular, it was necessary to keep the temperature and pressure constant for the full duration of a run. The main component of the experimental equipment was a 600 ml. stainless steel autoclave supplied by the Parr Instrument Company, 111. (Model 4560). The autoclave could be operated from room temperature up to 400°C at pressures ranging from atmospheric to 13.79 MPa (2000 psi). The principle features of the reactor (see Fig. 6.1) are summar ized below. 1. Variable speed stirrer, 0-600 r.p.m., driven by a drive assembly that could be easily disconnected and swung aside to allow full access to the autoclave head fittings. 2. Valves for adding gas, removing gas, and withdrawing liquid samples during runs. 3. A 0-2000 psi pressure gauge, accurate to within ± 5 psi. 4. A safety rupture disc. 5. A close fitting, quartz fabric heating mantle in an insulated alum inium housing. The heater could be easily lowered from the autoclave 45 Figure 6-1 Sketch of 600 ml. autoclave 47 without disturbing the stirrer or affecting any of the head connec tions . 6. A J-type thermocouple in a stainless steel well placed within the autoclave for measuring the reaction temperature. 7. An automatic temperature controller (Parr, Model 4831EB) whose output is monitored by a digital thermometer (Doric, Series 400A, Model 410A) and recorded on a strip chart recorder (Corning, Model 840). The controller was capable of holding the temperature to within ± 0.5°C of the desired value for an indefinite period (experiments lasting up to 30 days were performed). 8. An internal cooling coil, which was useful for controlling exothermic reactions and for rapid cooling at the end of a run. 9. The autoclave could be fitted with a pyrex liner so that experiments could be performed without the reactants coming into contact with the metal surface of the autoclave. 6.2 Loading the autoclave It was desirable to inject the aqueous DEA solution into the auto clave, which had been raised to the desired operating temperature. This measure minimized the problem of accounting for the time required to heat the autoclave and feed at the beginning of a run. A modified 500 ml. pressure sampling cylinder was used for the injection (see Fig. 6,2). The cylinder was first .purged with C02 or N2 depending on the type of run being conducted. The purging was necessary to remove any oxygen which could react with DEA forming heat stable salts. The cylinder was then filled with about 250 ml. of aqueous DEA solution and pressurized with C02 or N2 to the operating pressure. The contents of the cylinder were then discharged into the autoclave which was at operating Autoclave f DO—--KH—/ A 500ml. cylinder DEA CO2 cylinder Figure 6.2 Sketch of the autoclave loading system .0-00 49 temperature. A short amount of time (5-10 min.) was then required for the temperature and pressure to level off. 6.3 Sampling Sampling was done at the operating temperature and pressure by means of a 5 ml. coiled sample tube fitted with an inlet and outlet valve. The sample tube could be easily fitted and removed from the liquid sample port of the autoclave during a run. The following method was used for obtaining a sample from the reac tion mixture. The sample tube was first connected to the autoclave sample port. The autoclave sample valve was opened with the sample tube inlet and outlet valves closed. The sample tube inlet valve was then opened and a liquid sample was forced into the tube under the reactor pressure. The outlet valve was then opened slightly to bleed off a little sample. All the valves were then shut and the sample tube removed and placed in water for rapid cooling. The sample was then removed from the tube and stored in a glass sample bottle under a blanket of nitrogen for later use. 6.4 Analysis of the liquid samples The samples for the runs were stored in 7 ml. glass bottles with screw tops. 1 pi. samples were then injected into the gas chromatograph under the conditions described in chapter 3. When the 30 minute anal ysis was complete the peak areas and retention times of the major peaks of the chromatogram were recorded. Using the calibration curves (see Figs. 3.3-3.6) the concentrations of DEA and its major degradation pro ducts could be determined. Using these data, curves of concentration versus time could be plotted and studied. 50 6.5 Analysis of the gas phase The degradation of DEA by C02 did not result in the production of measurable amounts of gaseous degradation products. Aqueous samples of degraded DEA, which had been removed from the autoclave at various times during a run, were analysed for carbon, hydrogen, oxygen, and nitroge The concentration of each element in the liquid phase did not change during the runs. This indicates that no gaseous products were formed in the DEA degradation. Therefore, analysis of the gas phase was con sidered unnecessary. 6.6 Experimental procedure A typical controlled degradation run involves reacting an aqueous DEA solution with C02 at a desired temperature and pressure for a specific length of time. Samples were removed at regular intervals throughout the run. The subsequent procedure was followed. 1. The autoclave was first sealed empty. 2. The autoclave and modified sampling cylinder were then purged with C02 . 3. The autoclave was heated to the desired operating temperature. 4. About 250 ml. of fresh aqueous DEA solution of known concentration was charged to the modified sampling cylinder. 5. The aqueous DEA was injected under pressure into the autoclave. The stirrer speed was set at about 150 r.p.m. and several minutes were allowed for the temperature and pressure to settle down. Initially the C02 was absorbed by the DEA and the autoclave pressure had to be checked regularly over the first half hour of the run. C02 was added when necessary to keep the operating pressure constant. Also, since the absorption of C02 is an exothermic reaction, water was passed through the cooling coil to maintain the operating temperature. After about 30 min. the temperature and pressure became steady and required no further attention. (Usually much less than half an hour was required, depending on operating temper ature, pressure, and DEA concentration.) 6. Samples were removed at regular intervals during the run using the sampling tube. After the sample had been transferred to the sampling bottle the sampling tube was thoroughly cleaned with distilled water and dried with air. 7. When the run was completed, the heating jacket of the autoclave was switched off and removed. Water was then passed through the cooling coil and the autoclave and contents rapidly cooled to room temperature. 8. When the autoclave was at room temperature the pressure was reduced to atmospheric and the autoclave opened. Once the contents had been removed the autoclave was thoroughly cleaned with distilled water. Most runs were conducted with 250 ml. of solution in the 600 ml. autoclave. This volume of solution was considered sufficient so that the removal of several samples did not have a significant effect on the reactant volume. Also the reactant volume was not too large to limit the availability of C02^. It was hoped that C02 would be in excess for all the runs. A pressure of 600 psi, used in most runs, was chosen 34 41 partly to compare the results of this work with other studies ' and partly because it is close to pressures used industrially. The duration of the experiment depended simply on how long it took for significant degradation to take place. 52 6.7 Maintenance and performance The main problem with the autoclave was gas leakage around the stirrer shaft. There is a short hose nipple, which can be used to monitor the packing gland to detect any leakage as the packing elements and stirrer shaft become worn. Repacking, therefore, had to be carried out periodic ally. The frequency with which the gland was repacked depended on oper ating temperature and pressure, nature of the reactants, and the state of repair of the other various elements of the stirrer assembly. How ever, as a rule, the packing elements were usually replaced after 200 to 300 hours of service. On several occasions the stirrer shaft, in the vicinity of the packing, became worn and had to be replaced. Other than' leakage, the autoclave was relatively trouble free. Perhaps the only other problem was cleaning. Usually flushing the equipment with water was sufficient. However, a slow build-up of a thick viscous residue occurred and, periodically, the whole equipment had to be dismantled and thoroughly cleaned using water. 6 .8 Sources of errors It is possible that a sample may have a composition slightly differ ent from the composition of the solution in the autoclave. When a sample is removed from the autoclave it undergoes temperature and pressure changes which may cause the equilibria set up in the bulk liquid to change. Slight errors could also occur in recording .the time of sample removal. However this error becomes insignificant for runs of over eight hours. The temperature controller operates using a simple on/off control. This caused the temperature to oscillate between ± 1°C at 205°C. Again this error is minimal and, due to the regular oscillation, is averaged out. 53 Other small errors occur in the reading of the pressure gauge and loss of C02 through sampling and minor leaks. CHAPTER 7 PRELIMINARY EXPERIMENTS AND DEVELOPMENT OF THE EXPERIMENTAL PROGRAMME 7.1 Use of high temperatures for the degradation runs Since the degradation reaction under plant operating conditions is extremely slow it was decided to conduct the majority of experiments in the temperature range of 175°-205°C. In this range it is possible to achieve significant degradation in a matter of hours rather than weeks. However, it had to be established that the degradation products produced at elevated temperatures were similar to those produced under plant conditions. This was done by first comparing the results of degrad ing 30 wt % DEA with 4137 kPa (600 psi) C02 at 205°C and 120°C (see runs 3 and 11, details of which can be found in Appendix B). Secondly, degraded samples of DEA produced in the laboratory were compared with samples obtained from industrial DEA gas treating units. 7.1.1 Temperature comparisons. Figure 7.1 shows the chromatogram -3 of a degraded solution of DEA whose initial concentration was 3 x 10 moles/cc (~30 wt %) and had been degraded with C02 at 205°C for one hour. Figure 7.2 shows the chromatogram of a similar sample which was contacted with C02 at 120°C for 150 hr. _3 In both cases the DEA had degraded from 3 * 10 moles/cc to about _3 2.1 x 10 moles/cc, and three main degradation products were formed 54 HEOD Figure 7.2 Chromatogram of a 30 wt % DEA solution degraded at 120°C under 4137 kPa C02 (time—150 hr.) 56 although in different amounts. These compounds were identified as HEOD, BHEP, and THEED. The reason for the difference in concentration of these compounds will be discussed in chapter 11. 7.1.2 Comparison with industrial samples. Figures 7.3 and 7.4 show chromatograms of DEA treating solutions supplied by Aquitaine Canada Ltd. and Hudson's Bay Oil and Gas Company Ltd. Peaks of HEOD, BHEP, and THEED are evident. A fourth peak is present in the industrial samples which is probably TEA, since industrial DEA solutions contain significant quantities of TEA compared with reagent grade DEA. From Figs. 7.1-7.4 it is seen that there are strong similarities between the composition of DEA solutions degraded under laboratory condi tions at high and low temperatures and those degraded under industrial conditions. It may, therefore, be concluded that DEA undergoes essen tially the same kind of reaction in each case. 7.1.3 Thermal degradation. It is known that DEA can undergo thermal decomposition at its boiling point.Therefore, it had to be determined whether thermal degradation was significant at 205°C. A simple test was conducted in which a 30 wt % solution of DEA was heated to 205°C for 8 hr. under 4137 kPa (600 psi) of nitrogen (run 53). No significant change in the DEA concentration or the formation of degradation compounds were noted. However, in a similar test which lasted 200 hours (run 54), measurable quantities of BHEP and THEED were produced. Since all the runs conducted at elevated temperatures lasted only eight hours, thermal degradation was not considered to be significant. 7.1.4 Justification for the use of elevated temperatures. Experi ments conducted at temperatures well above the operating temperature of an industrial gas treating unit is justified for the following reasons. Figure 7.3 Chromatogram of a degraded DEA solution from a gas sweetening unit operated by Aquitaine Canada Ltd. Figure 7.A Chromatogram of a degraded DEA solution from a gas sweetening unit operated by Hudson's Bay Oil and Gas Co. Ltd. 58 1. Elevated temperatures accelerate the DEA degradation and make it possible to complete tests in hours rather than weeks. 2. Degradation products formed at elevated temperatures are quite similar to those produced under plant conditions. 3. Temperatures experienced by DEA in operating plants may, at certain points, be much higher than expected. For example, the surface temperature of the heating tubes in the stripper reboiler can be considerably higher than the bulk DEA temperature. 4. Thermal degradation of DEA is not a problem at temperatures up to 205°C. 7.2 Effect of metal surfaces It has been reported that the presence of metal ions may affect degradation.^'^ Therefore, to determine the influence of the stainless steel surface of the autoclave, several runs were conducted using a pyrex liner in the autoclave. The results of these runs were compared with those from identical runs where the liner was absent. No significant change was noted and therefore subsequent runs did not use the pyrex liner. 7.3 Effect of stirrer speed and reactant volume Various stirrer speeds in the range of 10-150 r.p.m. and reactant volumes in the range of 50-450 ml were used. The purpose of these runs was to determine if the mass transfer of C02 from the atmosphere above the solution to the DEA was affected by stirrer speed and/or volume of available C02. ,No significant effects were noted. This was probably due to the fact that the degradation reaction is extremely slow in compari son to C02 dissolving into the DEA solution (see chapter 2). Hence the DEA solution is saturated with C02 before any significant degradation 59 takes place and changes such as stirrer speed, have little effect. 7.4 Reproducibility 7.4.1 Samples. Degraded samples of DEA were, in general, very stable at room temperature. Analysis of samples over a two year period were found to be virtually identical. This not only demonstrated the stability of the samples but also the reliability of the analytical pro cedure . 7.4.2 Runs. Several runs were repeated over a two year period and were found to agree well within ± 5%. For example, Table 7.1 shows the concentration of DEA versus time for three different runs where a -3 3 * 10 moles/cc DEA solution was degraded at 175°C under 4137 kPa (600 psi) C02. Run C was conducted two years after runs A and B. Table 7.1 Comparison of reproducibility of degradation runs Sample Hours Concentration of DEA moles/cc Run A Run B Run C 0 3.00 x 10~3 3.085 x io"3 3.1 x 10 1 2.67 2.85 2.76 2 2.36 2.39 2.43 3 1.988 2.13 2.19 4 1.84 1.87 1.9 5 1.58 1.63 1.7 6 1.44 1.45 1.45 7 1.24 1.23 1.3 8 1.08 1.17 1.12 -3 60 7.5 Experimental programme Having established a simple experimental procedure to observe the degradation of DEA under controlled conditions, it was attempted to devise an experimental programme which could produce sufficient information to develop a kinetic model for the reactions. Experiments were there fore conducted to observe the effect of temperature, total pressure and initial DEA concentration. Table 7.2 summarizes the runs carried out. Table 7.2 Summary of experiments performed with C02 at 4137 kPa (600 psi) Initial DEA Concentration Temperature °C wt % 250 220 205 195 185 175 162 150 145 140 120 90 100 * * x 80 x 60 XXX 40 x 30 xxxxxxxxxxxx 20 X X X X 15 X 10 XXX 5 x A subsequent series of experiments degraded 30 wt % DEA at 195°C under the following pressures of C02:- 6895(1000), 5516(800), 4137(600), 3448(500), 2758(400), 2067(300), 1517(220) kPa (psi). The results of all these sets of experiments are tabulated in Appendix B. It was later discovered that these experiments were not sufficient to explain fully the degradation mechanism although they could be used as the basis for developing a kinetic model of the reactions. Therefore, further runs were performed to study the degradation mechanism and are described in chapters 9 and 10. For this reason the chapter on experi mental results and discussion has been split into three chapters. Chap ter 8 discusses experiments designed to study the kinetics, chapter 9 discusses experiments designed to study the degradation mechanism, and chapter 10 discusses experiments designed to study the behaviour of the major degradation compounds and feed impurities. CHAPTER 8 RESULTS AND DISCUSSION OF EXPERIMENTS DESIGNED TO STUDY THE KINETICS OF THE DEGRADATION REACTION The results of the degradation runs are tabulated in Appendix B in the form of concentration of DEA and its main degradation products versus time. Figure 8.1 shows typical chromatograms of a DEA solution degraded under lab conditions (run 3) corresponding to samples taken at 2 hour intervals. Three major degradation peaks produced by BHEP, HEOD, and THEED are evident. The MEA peak is an impurity in the intial solution. The HEOD and THEED peaks increase sharply at first and then either remain constant in size or decrease. This suggests that both HEOD and THEED are intermediate degradation compounds. The DEA peak decreases progressively whereas the BHEP peak grows. It was observed that the DEA solution slowly darkened with the colour changing from a light yellow to a dark brown as degradation pro gressed. Also there was a change in odour with the solution becoming more pungent. Although many other degradation compounds were detected especially at high temperatures, it is believed that it was only necessary to record, in full, data on DEA, BHEP, HEOD, and THEED. This conclusion was reached because the other minor degradation compounds existed generally in very low concentrations and, probably, had little effect on the kinetic model. 62 63 DEA Time: A hr . Time: 6 hr Figure 8.1 Typical chromatograms of a 30 wt % DEA solution degraded at 205°C under 4137 kPa C02 64 8.1 Effect of temperature Figures 8.2 and 8.3 show the change of DEA concentration with time when a solution consisting initially of about 30 wt % DEA is subjected to C02 at a pressure of 4137 kPa (600 psi) and temperatures ranging from 90-250°C (runs 1-12). The results have been plotted on semi-logarithmic 34 41 scales. The reason for this is that earlier work ' suggested the initial degradation of DEA to be governed by a first order reaction. Hence, when C02 is in excess, DEA was thought to degrade according to the following equations :-DEA DE^ products [8.1] dl^Al . _knEA [DEAlt [8.2] where [DEA] = concentration of DEA at time t Jt t = time k„. = overall reaction rate constant DEA The integrated form of Eqn. 8.2 is:-kDEA t log [DEA]t = log [DEA]Q - [8.3] If the degradation of DEA is first order then a semi-logarithmic plot of [DEA] versus t should be a straight line. Examination of Figs. 8.2 and 8.3 indicates that the data fall on straight lines especially at low temperatures. However, at high tempera tures, the degradation rate slows perceptibly after about 5 hours. In an extended run (run 32, Fig. 10.2) where 30 wt % DEA was degraded under 4137 kPa (600 psi) C02 at 205°C for 50 hours, the DEA concentration approached a nearly constant value of about 2 wt %. This suggests that the reaction is more complicated than initially suspected. It is possible DEA CONCENTRATION (xlO 3moles/cc) J L DEA CONCENTRATION (xlO 3moles/cc) 99 that C02 may cease to be in excess as the run proceeds, or the degrada tion products themselves may inhibit the degradation or the degradation reactions are reversible. These possibilities will be discussed later in chapter 11. Since it was not clear whether the intial degradation was governed by a first order reaction, further confirmation was sought. Using 69 Levenspiel's technique , the reaction order was obtained from studying initial degradation rates. Assuming C02 is in excess the general degrada tion reaction is of the form:-kDEA a [DEA] products [8.4] Hence . . ^ [DEA]* [8.5] or log{- d ^EA]} = log{kDEA) + a log {[DEA]} [8.6Therefore, if the log of initial degradation rate is plotted against log of the initial concentration, then a straight line should be produced with a slope equivalent to the reaction order. The initial degradation rates were calculated from runs of varying initial DEA concentration (runs 13-29) at 205, 175, and 150°C. The results are tabulated in Tables 8.1 to 8.3. Figure 8.4 shows the corresponding plots. Table 8.1 Initial DEA concentration and initial DEA degradation rates at 205°C Initial Run No. Initial concentration moles/cc degradation rate moles/cc.hr. 13 10 x io"3 2.2 x io-3 14 8 2 15 6 1.4 16 4 1.18 3 3 0.85 17 2 0.5 19 1 0.26 Table 8.2 Initial DEA concentration degradation rates at 175°C and initial DEA Run No. Init ial concentration moles/cc Initial degradation rate moles/cc.hr. 21 10 x 10~3 11 x 10"4 22 6 6 6 3 3.8 23 2 2.2 24 1.5 1.65 25 1 0.91 69 Figure 8.4 Initial degradation rate as a function of initial DEA concentration and temperature 70 Table 8.3 Initial DEA concentration and initial DEA degradation rates at 150°C Initial Run No. Initial concentration moles/cc degradation rate moles/cc.hr. 26 6 x io-3 12.9 x 10"5 8 3 7.1 27 2 3.8 28 1.5 2.6 29 1 2.0 Using the method of least squares lines were fitted to the plots of log(- iLLEIi^!) versus log [DEA]. The slopes of the lines were found to range from 1.025 to 0.996. Thus it can be assumed the initial degrada tion reaction is first order. A further check is provided by an Arrhenius plot based on the rela-tionship:-kD£A = A exp. {-E/RT} [8.7] log kDEA = log A -2-^ • i [8.8] where A = frequency factor E = activation energy R = universal gas constant T = absolute temperature If the data, when plotted as log k^EA vs. 1/T, fall on a straight line then the first order hypothesis is confirmed. Figure 8.5 is an Arrhenius plot where k^EA is calculated from the initial slopes of the curves in Figs. 8.1 and 8.2. As can be seen, the plot is linear at temperatures 71 ranging from 90-170°C. However, at higher temperatures, there is a departure from the straight line behaviour, once again indicating that more complex reactions are taking place. It is likely that the first order behaviour is only apparent, i.e., there may be several consecutive reactions taking place which are affected differently by temperature. What can be concluded, however, is that the reaction is highly sensitive to temperature. The initial degradation rate increases by nearly a factor of 3000 as the temperature is raised from 90 to 205°C. 8.1.1 Degradation products. Figures 8.6 to 8.11 show plots of HEOD, THEED, and BHEP concentration versus time. Solid lines are drawn through the experimental data points to indicate any observable trends. 8.1.1.1 HEOD. It can be seen from Figs. 8.6 and 8.7 that there is a rapid production of HEOD which levels off. The initial rate in creases with temperature although the overall amount of HEOD produced decreases. HEOD does not appear, therefore, to act as an intermediate 34 of the type suggested by the mechanism of Polderman and Steele. It may be possible for HEOD to be a final product of DEA degradation, which is thermally unstable. This point will be discussed further in chapter 11. 8.1.1.2 THEED. Figs. 8.8 and 8.9 show that the THEED concentra tion increases with time at a slightly lower rate than the HEOD concen tration, reaching a maximum value before decreasing again. The time required to reach the maximum concentration decreases with increasing temperature. THEED appears to behave more like an intermediate than HEOD. Figure 8.6 HEOD concentration ns a function of time and temperature (30 wt % DF.A, 4137 kPa C02, 205-162°C) Figure 8.7 HEOD concentration as a function of time and temperature (30 wt % DEA, 4137 kPn C02, 150-90°C) TIME (hr) Figure 8.8 THEED concentr.ition as a function of time and temperature (30 wt % DEA, 4137 kPA C02, 205-l62°C) as TIME (hr) Figure 8.10 BHEP concentration as a function of time and temperature (30 wt % DEA, 4137 kPn CO,, 206-162°C) —i i 1 1 1 \ 0 20 40 60 80 100 120 140 160 TIME (hr) Figure 8.11 BHEP concentration as n function of time and temperature (30 wt % DEA, 4137 kl'n CO, , 150- 90°C) 79 8.1.1.3 BHEP. Figures 8.10 and 8.11 show that the concentration of BHEP rises steadily with time. The overall production of BHEP increas ing rapidly with temperature. At temperatures greater than about 185°C, the production of BHEP (i.e., the slope of the BHEP concentration versus time curve) starts to fall slightly after several hours and this implies that the concentration of a certain intermediate is falling. HEOD cannot be this intermediate since its concentration remains relatively constant after an initial period and this would result in the production of BHEP becoming constant. THEED is more likely to be the intermediate respon sible for the formation of BHEP. THEED's concentration falls after reaching a maximum and this would cause the production of BHEP to fall as observed in Figs. 8.10 and 8.11. This will be discussed further in chapter 11. 8.2 Effect of initial DEA concentration The results of degradation experiments conducted by using different initial DEA concentration were rather confusing. Figures 8.12 to 8.14 show the change in DEA concentration with time for varying solution strengths at temperatures of 205°C, 175°C, and 150°C. Again, the plots tend to deviate from the straight line behaviour at high temperatures. If it is assumed that the initial degradation rate is first order, then kDEA should be independent of the initial DEA concentration. How ever this is clearly not the case as shown by Fig. 8.15, which is a plot of kp£A versus initial DEA concentration. It appears there are three regimes:-1. 0-10 wt % where the degradation rate constant of DEA is constant at a low value. 2. 10-30 wt % where the rate constant rapidly increases with increasing TIME (hr) Figure 8.12 DEA concentration ns a function of time and initial DEA concentration (4137 kPn C02, 205°C) 00 o TIME (hr) Figure 8.13 DEA concentration as a function of time and initial DEA concentration (4137 kPa CO,, 175°C) t—* 10 20 30 40 50 TIME (hr) Figure 8.14 DEA concentration ,is a function of time and initial DEA concentrati (4137 kPa CO,, 150"C) INITIAL DEA CONCENTRATION (wt %) Figure 8.15 k^^ as a function of initial DEA concentration and temperature (4137 kPa C02) 84 initial DEA concentration. 3. 30-100 wt % where the rate constant is high and relatively constant, decreasing slightly as the initial DEA concentration nears 100 wt %. Figure 8.16 shows the Arrhenius plot for various initial DEA con centrations. Again the three regions can clearly be seen. At this stage it was not possible to explain this behaviour. Thus more experiments had to be conducted and these are reported in chapters 9 and 10. Figures 8.17 to 8.19 show typical plots of HEOD, THEED, and BHEP concentrations versus time as a function of initial DEA concentration at 205°C. As before the HEOD concentration rises rapidly and then levels off, whereas THEED tends to a maximum concentration before falling again (especially at high initial DEA concentrations). BHEP once again appears to be produced from THEED rather than HEOD. It is interesting to note that the production of BHEP and HEOD are both lower at an initial DEA concentration of 100 wt % than at 80 wt %. This is reflected in the kDEA values being lower at 100 wt % than 80 wt %. This behaviour indicates that water may play a significant role in the overall degradation. Initially, at 100 wt % DEA, the only water present is in the form of a trace impurity in the DEA feed. As the degradation reaction proceeds water is produced as a degradation product (see chapter 2, Eq. 2.17). 8.3 Effect of pressure Typical plots of changes in DEA concentration versus time as a function of overall pressure are shown in Figure 8.20. These experiments were conducted at 195°C. At this temperature the water in the 30 wt % DEA solution exerts a considerable pressure of about 1202 kPa (174.3 psi). Therefore, it must be noted that the partial pressure of C02 85 i 1 r I i I I I 1 L_ 2.1 2.2 2.3 2.4 2.5 2.6 . 1000/T (°k"') Figure 8.16 Arrhenius plot for varying DEA solution strengths degraded with C02 at 4137 kPa Figure 8.18 THEED concentration as a function of time and initial DEA concentration . (4137 kPa C02, 205°C) 1 1. 0 12 3 4 5 6 7 TIME (hr) Figure 8.19 BHEP concentration ns ;i function of time and initial DEA concentration (4137 kPa C02, 205°C) 00 00 is considerably less than the total pressure. From Fig. 8.20 it can be seen that degradation increases with increasing pressure up to about 4137 kPa (600 psi) total pressure; above that pressure little change is noted. Figure 8.21 shows the initial k^^ values as a function of total pressure when a solution containing initially 30 wt % DEA is de graded at 195°C. Both Figs. 8.20 and 8.21 imply that C02 is limiting at pressures below 4137 kPa (600 psi). Therefore, it was necessary to determine the solubility of C02 in DEA solutions at these overall pres sures. Using these concentrations it could then be determined whether or not industrial units were operating under C02 limiting conditions and how to relate the results of this study to industrial units. Unfortu nately, high temperature data on C02 solubility in DEA solutions were not available in the open literature. Hence solubility experiments were performed to obtain these data and are discussed in chapter 9 and Appendix C. Figures 8.22 to 8.24 show plots of HEOD, THEED, and BHEP concentra tion versus time as a function of overall reaction pressure at 195°C. Figure 8.20 DEA concentration as a function of time and pressure (30 wt % DEA, 195°C) (PSI) 00 400 600 800 1000 T I 1 1 1— REACTION PRESSURE (kPa) Figure 8.21 kp,-. as a function of DEA (30 wt % DEA, 195°C) reaction pressure Figure 8.22 HEOD concentration as a function of time and pressure (30 wt % DEA, 195°C) IO TIME (hr) Figure 8.23 THEED concentration as a function of time and pressure (30 wt % DEA, 1.95°C) BHEP CONCENTRATION (xlO ''moles/cc ) TO C i-l ro o rr > 03 rq -o o o 3 o n 3 rr "1 Cu n 3 n o 3 3 ro CJ 3 C •a cn •J: C -1 CHAPTER 9 EXPERIMENTS DESIGNED TO ELUCIDATE THE DEGRADATION MECHANISM Before a kinetic model could be devised it was necessary to gain insight into the degradation mechanism. The experiments described in chapters 7 and 8 gave some idea of the mechanism but they also generated many questions which needed to be answered such as, "why does the initial DEA concentration affect the reaction rate?" and "why do the plots of log [DEA] vs. time deviate from the linear behaviour at high temperatures?" Therefore, the following studies were conducted in order to explain the questions posed by chapters 7 and 8 and to develop a comprehensive model of the degradation of DEA. 9.1 Effect of pH Since the DEA solutions are complex mixtures of ionized species, it is highly likely that changes in pH will affect the equilibrium and, in turn, the overall degradation reaction.^'^. DEA itself tends to make the solution alkaline whereas the dissolved C02 tends to render the solution acidic. Runs were carried out where the DEA feed was made more alkaline or acidic by adding NaOH or HCL respectively (runs Al to 43). The effect of these changes in pH can be seen in Fig. 9.1 where DEA concentration is plotted as a function of time for different initial solution pH measured at room temperature and at atmospheric pressure. It must be realized that the pH of the reaction mixture under operating 95 TIME (hr) Figure 9.1 DEA concentrn t ion as a function of time and solution pH (30 wt % DEA, A137 kPa C02 , 205°C) conditions will be different to that at room temperature. However, it was not possible to measure the solution pH under operating conditions with the equipment available. An aqueous solution containing 30 wt % DEA has a pH of about 11.2 at room temperature. Therefore, the run at a pH of 11.2 in Fig. 9.1 (run 3) can be considered the reference exper iment . As seen from Fig. 9.1, lowering the pH reduces the degradation rate. By changing the initial solution pH from 12.24 to 9 the degrada tion rate is reduced by a factor of over 5. The effect of pH can be linked with the solubility of C02 in the DEA solution. The solubility of C02 being increased by the action of hydroxyl ions:-C02 + 0H~ ^ HC03~ [9.1] At low pH the solubility of C02 is greatly reduced and hence the degrada tion rate is lower. Further studies revealed that, when NaOH is added to HEOD in solu tion most of the HEOD is converted to DEA. This indicates that the HEOD ring is unstable. It appears that the electron deficient carbonyl atom of the ring is attacked by OH , resulting in ring opening.^ 0 II 0 CjHit-OH HEOD DEA At present, it is not possible to explain why a reduction in HEOD concentration increases DEA degradation. This will be investigated further in chapter 11. 98 9.2 Effect of bicarbonate and carbonate ions DEA treating solutions with dissolved C02 contain various ionic and molecular compounds such as: R2NH, R2NH2 + , R2NC00~, HC03~' C03 , and C02. Since the pH has such a strong effect on degradation, it is likely that degradation involves some ionic compounds. Therefore, tests were conducted where DEA was reacted with potassium carbonate (K2C03) and potassium bicarbonate (KHC03) under 4137 kPa (600 psi) of nitrogen. Thus initially the reaction mixture contained C03 or HC03 and no free C02. Under these highly alkaline conditions it is virtually impossible for HC03 or C03 to directly revert to free C02. For these runs the molar concentrations of C03 and HC03 were made equivalent to that of C02 dissolved in DEA under normal reaction conditions. This was to ensure that any changes noted in the degradation reaction were not due to differences in the amount of C02 in the reaction mixture either as free C02 or HC03 or C03 . Unfortunately the open literature did not provide data on C02 solubility in DEA under the reaction conditions of this study. Therefore, a series of solubility experiments had to be performed to produce the necessary data. 9.2.1 CO; solubility data. A simple method for determining the solubility of C02 in DEA solutions under high temperature and pressure was developed. Details of the method used are given in Appendix C. Solubilities were determined for the following conditions:-DEA concentration - 30, 20, and 10 wt % Temperature - 205 to 100°C Overall pressure - 413.7 to 4137 kPa (60 to 600 psi) It must be noted that the overall pressure in the autoclave was made up from C02 and water vapour from the aqueous DEA solution (and 99 to a small extent from DEA itself). For example at 205°C the vapour pressure of a 30 wt % DEA solution is 1503 kPa (218 psi). Figures 9.2 to 9.4 summarize the results of the solubility experiments. It is realized that the method used for determining the C02 solubil ity was very simple and probably not very accurate. However the purpose of these experiments was only to obtain approximate solubilities to within ± 10%. To check the accuracy of the method, the solubility results were compared to those in the literature where the data overlapped. The comparisons are shown in Table 9.1. Table 9.1 Comparison of C02 solubilities in DEA solutions DEA Cone wt % Temp °C C02 partial pressure psi kPa C02 concentration g C02/g DEA This study Literature % Difference Reference 20 100 100.0 689.5 0.272 0.27 0.74 71 20 120 100.0 689.5 0.238 0.212 12.26 71 20 140 100.0 689.5 0.204 0.186 7.53 71 20 100 316.0 2178.8 0.366 0.348 5.17 71 20 120 316.0 2178.8 0.331 0.294 12.58 71 20 140 316.0 2178.8 0.29 0.25 16.00 71 25 107 85.8 591.6 0.22* 0.218 0.92 72 25 107 317.0 2185.7 0.31* 0.287 8.01 72 25 121 48.7 335.8 0.165* 0.143 15.38 72 25 121 230.0 1585.9 0.275* 0.248 10.90 72 25 121 402.0 2771.2 0.33* 0.3 10.00 72 Av = 9.04 *Extrapolated results from Figs. 9.2 and 9.3 PARTIAL PRESSURE C02 (PSI) PARTIAL PRESSURE C02 (PSI) 40 50 60 80 100 200 300 400 500 n i i i i 1 1 1 r (kPa) Figure 9.3 Solubility of C02 in 20 wt % DEA 10 0.6 0.5 20 PARTIAL PRESSURE C02 (PSI) (kPa) Figure 9.4 Solubility of C02 in 10 wt % DEA o ho 103 9.2.2 Runs using HC03~ and C03" instead of C02 The runs performed are summarized in Table 9.2. Table 9. 2 Summary of HCO," and C03~ runs Run No. DEA Cone wt % Temp °C Nitrogen psi Pressure kPa Ion* Run time hr. 44 30 205 600 4137 co3" 8 45 30 205 600 4137 HC03 30 46 30 175 600 4137 HC03~ 24 47 30 150 600 4137 HC03~ 50 *The amount of the salt used was calculated by the following method: For example using run 41 the solubility of C02 under operating con ditions £ 0.17 g C02/g DEA Wt. of DEA solution = 300 g MW of C02 = 44 Cone, of DEA = 30 wt % MW of KHC03 = 100 Weight of DEA = 90 g Required C02 = 90 x 0.17 = 15.3 g Required KHC03 = 15.3 * 100/44 = 34.8 g Run 44 using K2C03 resulted in no degradation. Therefore C03 can be considered to play no part in the degradation of DEA. Since the solution was highly alkaline due to the presence of both DEA and HC03 , it was assumed that C02 was only present in the carbonate form. Figure 9.5 shows DEA concentration as a function of time for runs 45 to 47 in which KHC03 was used. The plots are linear up to about -3 0.9 x 10 moles/cc DEA. Figures 9.6 to 9.8 show the corresponding plots for the degradation products. The degradation appears similar to that observed by using pure C02 but the rates are very much lower. For example, the degradation products formed at 205°C (run 45) are pro duced in similar amounts to an equivalent run using C02 (run 3, Fig. 8.2) instead of KHCO, . Table 9.3 shows the k^,,. values for the KHCO, 3 DEA 3 10 20 30 TIME (hr) Figure 9.5 DEA concentration as a function of time and temperature (using KHCO,, 30 wt % DEA, 4 137 kPa N2) TIME (hr) Figure 9.6 HEOD concentrntion as a function of time and temperature (using KHCO, , 30 wt 7. DEA, 4137 kPa N2) i—• o Figure 9.7 THEED concentration as a function of time and temperature (using KHCO,, 30 wt % DEA, 4137 kPa N2) LOl 108 runs and equivalent CO, runs. Also, the k__. values for runs where n 2 DEA 10 wt % DEA is degraded under C02 are listed. The close similarity between the k^_, values for the ionic runs and the CO, runs at 10 wt DEA 2 % DEA is noteworthy. Table 9.3 Comparison of k^^. values for runs conducted with KHC03 and C02 ^EA (hr_1) Temp 30 wt % ionic 30 wt % standard 10 wt % standard °C (KHC03 Runs) (C02 Runs) (C02 Runs) 205 0.104 0.29 0.101 175 0.026 0.121 0.0242 150 0.0053 0.031 0.0055 The major difference between the ionic runs and the standard C02 runs is the production of HEOD. Very little HEOD is produced using KHC03. The lower production of HEOD could be linked with the fact that the solution is more basic than under standard conditions (KHC03 being more basic than DEA). As mentioned previously, increasing the alkalinity causes HEOD to break down and to inhibit the production of HEOD. From these results it can be concluded that DEA degradation can be caused by HC03 . However, other reactions must also take place since the degradation increases when pure C02 is used. 9.3 Effect of water To investigate degradation by other than ionic routes, degradation experiments were conducted with C02 in the absence of water. This was achieved by diluting DEA with methyldiethanolamine, MDEA. MDEA is similar to DEA but relatively inert to C02 and does not react with DEA under 109 operating conditions (see run 53). Since water was absent from the reaction mixture, ions could not be formed and hence degradation could only be caused by C02 reacting directly with DEA. Since no ions are present, these runs are called "molecular runs". It must be remembered that the reactions producing HEOD, THEED, and BHEP all are accompanied by the production of water. However water is never in excess as in the standard runs performed with aqueous DEA solutions. The runs per formed are summarized in Table 9.A. Table 9.4 Summary of the molecular runs DEA cone Temp Pressure of C02 Run No wt % °C psi kPa 48 66.7 205 600 4137 49 40 205 600 4137 50 30 205 600 4137 51 30 175 600 4137 52 30 150 600 4137 Figures 9.9 to 9.12 show the results of the runs at 205=C. Once again the three major degradation products are formed in relatively the same amounts as in the standard runs; however, the rate is slower and k^ decreases slightly with initial DEA concentration. Thus it appears that DEA can degrade by two parallel reaction paths, one involving pure C02 and the other HC03 . Table 9.5 gives the values of kpEA for the molecular runs and compares them with those for the ionic runs and standard runs. Figure 9.10 HEOD concentration as a function of time and initial DEA concentration (no water present, 4137 kPn C02, 205°C) INIT. DEA CONC. TIME (hr) Figure 9.11 THEED concentration as a function of time and initial DEA concentration (no water present, 4137 kPa C02, 205°C) Figure 9.12 BHEP concentration as a function of time and initial DEA concentration (no water present, 4 137 kPa C02, 205°C) 114 Table 9.5 Comparison of k^^ for molecular, ionic, and standard runs DEA cone Temp ^EA ^hr ^ wt % °C Molecular Ionic Standard Molecular + Ionic 100 205 0.195 - 0.195 -66.7 205 0.175 - 0.3 -40 205 0.168 - 0.32 -30 205 0.14 0.104 0.29 0.244 30 175 0.075 0.025 0.121 0.10 30 150 0.0203 0.0053 0.031 0.0253 Assuming there are two parallel reactions degrading DEA then basic kinetic theory indicates that the overall degradation rate is the sum of the rates for the two parallel reactions, i.e., (kn_.) „ = (kn„.). . + (k__A) . , [9.3] DEA overall DEA ionic DEA molecular Referring to Table 9.5, it can be seen that the sum of the k values for the two degradation routes is close to the k values for the standard run. In all cases the sum the rate constants is lower than the standard value and this is probably due to the fact that the molecular runs exclude water. In a normal run water is always present and it seems likely that water may help the molecular route hence increasing kp£A for the molecular runs. Table 9.5 also shows that the k^. for the molecular DEA runs decreases slightly with decreasing concentration. A possible reason is that another degradation product is water. At high concentrations of DEA, more water is produced which can aid the overall degradation. The implications of the possibility of two degradation routes will be discussed further in chapter 11 and an attempt will be made to explain why the initial DEA concentration affects the overall rate constant k^^A (see Fig. 8.15). 115 9.A Thermal degradation One of the reasons for the difficulty in purifying degraded DEA solutions is the fact that DEA breaks down at its boiling point. A simple test where DEA was boiled at atmospheric pressure under nitrogen for about 30 minutes resulted in 20% degradation producing THEED and BHEP. It was, therefore, decided to investigate the thermal degradation of 30 wt % DEA under operating conditions. Two runs were carried out under conditions summarized in Table 9.6. Table 9. .6 Summary of thermal runs Run No Initial DEA Cone wt % Pressure psi (Nitrogen) kPa Temp °C Duration of run hr. 5A 30 600 A137 205 200 55 30 600 A137 250 25 Figure 9.13 shows plots of DEA concentration as a function of time. They are both linear on semi-logarithmic scales indicating a first order reaction with the reaction rate being about one hundredth of that under standard conditions at 205°C (see Table 9.7). Table 9.7 Comparison of kp^^ for thermal and standard runs DEA cone wt % Temp °C ^EA Thermal (hr"1) Standard 30 205 0.00365 0.29 • 30 250 0.033 0.69 911 117 BHEP and THEED were the major degradation products. Figures 9.14 and 9.15 show plots of concentration versus time. The plots tend to indicate that a series reaction is taking place, i.e., DEA • THEED —• BHEP [9.4] The second reaction is more temperature sensitive than the first. Although the thermal degradation of DEA appears first order, it does not agree with the stochiometric equations, where two molecules of DEA are required to produce one molecule of THEED or BHEP. Therefore the reaction is not as simple as it appears. It is possible that an inter mediate is slowly produced from DEA via a first order reaction, which is rapidly consumed to produce THEED. For example:-DEA • Intermediate [9.5] 2 Intermediate • THEED + H20 [9.6] THEED • BHEP + H2 0 [9.7] What the results show is that BHEP can be produced from THEED which 34 does not agree with the mechanism proposed by Polderman and Steele. Hence, thermal degradation represents a third route for the degradation of DEA although its contribution is insignificant at low temperatures. The major degradation products BHEP and THEED are produced in similar amounts in the thermal runs under nitrogen to the standard runs under C02 although at a vastly decreased rate. It is possible, therefore, that C02 is just acting as a catalyst. In a typical run and even for runs of over 200 hr (e.g., run 10) C02 is neither produced nor consumed which tends to confirm the possibility that C02 is acting as a catalyst. TIME (hr) Figure 9.15 BHEP concentration as a function of time and temperature (no C02 present, 30 wt % DEA, 4137 kPa N2) 120 9.5 C02 solubility studies It was assumed that C02 would be in excess for all the experiments and therefore could be neglected in devising a kinetic model. However, referring to Fig. 8.21, chapter 8, it was noted that the total reaction pressure affected the degradation rate up to about 4137 kPa (600 psi). Using Fig. 9.2 it was possible to plot k^EA as a function of C02 concen tration in DEA (see Table 9.8 and Fig. 9.16). Table 9.8 Overall kD£A as a function of C02 solubility for degradation runs of 30 wt % DEA at 195°C Total psi Pressure kPa Partial psi Pressure of C02 kPa Solubility g C02/g DEA kDEA hr"1 1000 6895 825.7 5693.2 0.312 0.23 800 5516 625.7 4314.2 0.259 0.23 600 4137 425.7 2935.2 0.196 0.23 500 3448 325.7 2245.7 0.165 0.185 400 2758 225.7 1556.2 0.127 0.154 300 2069 125.7 866.7 0.08 0.098 220 1517 45.7 315.1 0.043 0.061 From Fig. 9.16 it appears that if the C02 concentration falls below approximately 0.2 g C02/g DEA, then C02 becomes limiting and must be included in the degradation model. It was found that at high temperatures the concentration of C02 is very close to 0.2 g C02/g DEA in a 30 wt % DEA solution under 600 psi C02. At 195°C the C02 concentration falls below the 0.2 level (see Table 9.9). SOLUBILITY OF CO2 IN DEA (g C02/g DEA) Figure.9.16 kDEA as a function of C02 concentration (30 wt % DEA, 195°C) 122 Table 9.9 C02 solubility as a function of temperature in a 30 wt % DEA solution under a total pressure of 4137 kPa (600 psi) Temperature Partial Pressure of C02 Solubility of C02 °C psi kPa g C02/g DEA 205 381.4 2629.8 0.168 195 425.7 2935.2 0.194 185 461.4 3181.4 0.221 175 494.1 3406.8 0.245 150 544.3 3752.9 0.32 120 575.4 3967.4 0.38 Also increasing the concentration of DEA tends to reduce the ratio of C02 to DEA as shown in Table 9.10. Table 9.10 C02 solubility as a function of DEA concentration for solutions at 205°C under a total pressure of 4137 kPa (600 psi) Concentration of DEA Partial pressure of C02 Solubility of C02 wt % psi kPa g C02/g DEA 30 381.4 2629.8 0.168 20 370.0 2551.2 0.196 10 358.0 2468.4 0.28 Therefore, it can be concluded that for runs at high DEA concentra tion and high temperatures, the concentration of C02 will be very close to if not below the value of 0.2 g C02/g DEA. Thus a possible explanation for the deviation from the straight line behaviour observed in certain DEA concentration versus time plots could be due to changes in the C02 concentration in the reaction mixture. Although it appears that no C02 is consumed during a reaction (i.e., there is no change in pressure during a run) it is possible that C02 is being converted to a form which reacts slower with DEA (i.e., see the ionic runs discussed in section 9.1) or the C02 becomes tied up in some manner with the degradation com pounds (see the BHEP runs discussed in section 10.2). Referring to Fig. 8.20, it can be seen that, as the pressure is increased beyond 4137 kPa (600 psi) the deviation of the plot from the linear form decreases, although the initial k^. remains the same. This indicates, for example, that for the 6895 kPa (1000 psi) run (run 33) the concentration remains above 0.2 g C02/g DEA for the entire run. A run (run 55) was performed where 30 wt % DEA was degraded at 205°C under twice the usual C02 pressure, i.e., 8275 kPa (1200 psi). Figure 9.17 shows a comparison of the results of the run at 8275 kPa (1200 psi) (run 56) with a standard run at 4137 kPa (600 psi) (run 3). The plot of log [DEA] versus time is completely linear at 8275 kPa (1200 psi) whereas the standard run starts to deviate after only 2 hours of reaction time. This clearly demonstrates that, at high temperatures, C02 limitation affects the rate of degradation. This will be further discussed in chapter 11. Figure 9.17 DEA concentration as a function of time and C02 pressure (30 wt % DEA, 205°C) CHAPTER 10 EXPERIMENTS DESIGNED TO STUDY THE BEHAVIOUR OF THE MAJOR DEGRADATION COMPOUNDS AND IMPURITIES IN THE DEA FEED In general, the degradation of DEA and production of its degradation compounds can be summarized by the qualitative plots shown in Fig. 10.1. The plots suggest that BHEP is produced in a series reaction from DEA via THEED. This hypothesis needed to be confirmed and, also the role of HEOD needed to be understood. Furthermore, it was necessary to deter mine whether any equilibrium reactions played a role in DEA degradation. To answer these questions, the following tests were performed (see Table 10.1). 10.1 Long term run Figure 10.2 shows a plot of concentration versus time for DEA, HEOD, THEED, and BHEP. As can be seen, DEA, HEOD, and THEED all tend to zero. BHEP appears to be the main degradation compound. However, under these extreme conditions many other compounds are produced and BHEP accounts for only about 50% of the DEA lost. It is possible that under these conditions high molecular weight polymers are produced which may not be detected by gas chromatography. The conclusion from this run is simple, overall there is no equilibrium between DEA and its degra dation products. ' 125 126 TIME Figure 10.1 Typical plots of concentration as a function of time for DEA and its degradation products TIME (hr) No Figure 10.2 Concentration of DEA, HEOD, THEED, and BHEP as a function of time ^ (30 wt % DEA, 4137 kPa C02, 205°C) 128 Table 10.1 Summary of runs to study the behaviour of the major degradation compounds Feed Concentration Run lun moles/cc Temp Time lo. DEA HEOD THEED BHEP °C Gas hr. Comments 33 3*10~3 - - - 205 C02 50 long term run 57 _ _ -4 5x10 205 C02 8 BHEP studies 58 -3 3*10 _ — -4 5*10 205 N2 8 11 59 3*10"3 2.7*10~4 3. -4 .3 xlO 3.55xl0-4 205 C02 8 ti 60 3*10"3 _ — 4.7xl0_A 205 C02 50 11 61 3*10~3 - - 4.7xl0~4 150 C02 60 11 62 _ 3*10~3 - - 205 N2 1 HEOD studies 63 1.0*10" •3 4.24*10~4 - - 205 N2 8 it 64 1.5*10" •3 -4 4x10 9 -4 .7x10 0.28*10~4 175 C02 8 tt 65 1.5x10" -3 -4 4x10 9 -4 .7x10 0.28xl0"4 175 N2 8 it 66 — - 2 .6xl0~3 - 205 co2 1 THEED studies 67 _ 2 .6*10~3 - 205 N2 1 11 68 1.2x10* -3 - 2 .6xl0~3 205 N2 8 ft 69 1.2x10" -3 - 2 .6xl0~3 - 205 co2 8 It All the runs were performed at 4137 kPa (600 psi) 10.2 BHEP runs The stability of BHEP was first tested in run 57 and it was found that BHEP did not undergo any form of degradation. In run 58 BHEP was mixed with DEA and pressured to 4137 kPa (600 psi) of nitrogen to see whether BHEP would react with DEA. Again, no reaction was observed to take place. Thus it can be concluded that BHEP is a final degrada tion product and that, overall, DEA is slowly converted to BHEP by a slow complex series of degradation reactions. Run 59 was carried out to determine whether the presence of BHEP and other degradation products had any effect on the overall degradatio 129 The feed was made up from a mixture of degraded DEA solution and fresh DEA to give an overall DEA concentration of about 30 wt %. A similar run was performed where 30 wt % DEA was degraded in the presence of only BHEP (run 60). . Figure 10.3 compares the degradation of DEA for runs 59 and 60 with a standard run (run 3). As can be seen, the DEA degrada tion appears to be inhibited by the presence of degradation products, especially BHEP. A possible explanation for this behaviour is that degradation products (and/or BHEP) are tying up some of the available C02 dissolved in the reaction mixture. This may then cause the ratio of available C02 to DEA to fall below 0.2 g C02/g DEA and therefore reduce the rate of degradation. Run 60 was extended to 50 hours after which the reaction mixture was virtually identical to that of the standard long term run (run 33). In a similar experiment to run 60 where the reaction was performed at 150°C instead of 205°C (see run 61) no inhibiting effect due to BHEP was observed. Thus it appears as if the presence of BHEP or degradation products tends to slow down the rate of degradation only in situations where the concentration of C02 is very close to or below the limiting value of 0.2 g C02/g DEA. Also run 60 shows that the addition of BHEP to the reaction mixture has no effect on the overall production of BHEP from DEA. Therefore, it can be concluded that BHEP is not in equilibrium with DEA or other degradation products. 10.3 HEOD runs An aqueous solution of HEOD was heated to 205°C for one hour under nitrogen (run 62). DEA, THEED, and a trace of BHEP were produced. This indicated that there must be some form of equilibrium between HEOD and DEA. It is unlikely that there is any equilibrium between HEOD and u TIME (hr) Figure 10.3 DEA concentration as a function of time and degradation products (30 wt % DEA, 4137 kPa C02, 205°C) 131 THEED since THEED is unable to form HEOD (see run 66 or 67). However, it was not possible at this stage to confirm whether THEED and BHEP were being produced from HEOD or DEA. Figure 10.4 shows the concentration versus time curves for the degradation of a mixture of HEOD and DEA under nitrogen (run 63). As can be seen, the HEOD is mainly converted to THEED and a trace of BHEP; the DEA loss is fairly small. The feed used in runs 64 and 65 were produced by degrading a 30 wt % DEA solution at 175°C under 4137 kPa (600 psi) of C02 for 6 hr. The product was removed from the autoclave and heated to drive off any dissolved C02. Run 64 is an extended standard run (see run 6) and, in run 65 the mixture is degraded under nitrogen instead of C02. Figures 10.5 to 10.8 compare the results of the two runs. For run 64 it appears that HEOD is playing virtually no part in the degradation of DEA and its con centration remains nearly unchanged. THEED increases then falls slightly whereas the BHEP concentration increases steadily. When the same reaction is carried out under nitrogen (run 65) some major differences are noted. The rate of DEA degradation is less than that of the run under C02, but it is still quite significant. Since it was established earlier (section 9.4) that DEA does not degrade noticeably at 175°C under nitrogen, it is clear that the C02 and/or HC03 are provided either by the breakdown of HEOD to DEA or the formation of THEED. It can be seen from Fig. 10.6 that HEOD does not break down completely but its concentration levels off after sharply falling. This indicates some form of equilibrium has been established between HEOD and its breakdown products or DEA. TIME (hr) 10.4 Concentration of DEA, HEOD, THEED, and BHEP as a function of time (reactants—DEA and HEOD, 4137 kPa N2» 205°C) TIME (hr) Figure 10.5 DEA concentration as a function of time (reactants—degraded DEA solution, 4137 kPa C02 or N2, 175°C) LO u o E o z o I— < »— z LU u Z o u a O 'A— 2.0 1.0 RUN n A— 64 (COjl O—65 IN J 8 TIME (hr) Figure 10.6 HEOD concentration as a function of time (reactants—degraded DEA solution, A 137 kPa C02 or N2, 175°C) CO u o-UJ I I I I I I I _l L_l 0 12 3 4 5 6 7 8 TIME (hr) OJ Figure 10.8 BHEP concentration as a function of time °^ (reactants—degraded DEA solution, 4137 kPa C02 or N2, 175°C) 137 What appears to be taking place is that HEOD reacts to form either DEA or THEED and HC03 . The bicarbonate ion then reacts with DEA to produce either more HEOD or other degradation products. In the case of run 65 it seems that DEA just degrades to THEED and BHEP once sufficient HC03 has been produced from the breakdown of HEOD. In the case of run 64 it appears that HEOD need not undergo any breakdown since the solution has sufficient HC03 or C02 provided by the C02 atmosphere. In summary the following points can be made:-1. DEA can degrade to HEOD in the presence of C02. 2. HEOD can break down to DEA. 3. HEOD does not break down in the presence of DEA and C02• 4. HEOD partially breaks down in the presence of DEA and N2. Points 1 to 4 imply that HEOD may be involved in a reversible reaction with DEA. However, it is not a straightforward equilibrium and it is possible that HEOD is in equilibrium with a degradation product of DEA which is in turn in equilibrium with DEA. This intermediate product, if it exists, is not detected by gas chromatography or is not produced in sufficient quantities to measure. 10.4 THEED runs An aqueous solution of THEED was heated to 205°C under 4137 kPa C02 for one hour (run 66). The only product produced was BHEP. A similar run under 4137 kPa (600 psi) nitrogen (run 67) also produced BHEP but in a much smaller quantity. It, therefore, appears that BHEP can be directly produced from THEED with C02 acting like a catalyst. Runs 68 and 69 were conducted to determine the effect of DEA on the reac tion of THEED. Figure 10.9 shows the plots of concentration versus time for run 68. The concentration of DEA remains unchanged for the CONCENTRATION (xlO 3moles/cc) 139 8 hour run. This indicates that DEA and THEED do not react. Figure 10.9 also shows that, as the THEED concentration decreases the BHEP con centration increases proportionately. This indicates a stochiometric relationship between the two compounds. Figure 10.10 shows the effect of C02 which speeds up the conversion of THEED to BHEP (run 69). DEA degrades slightly due to the presence of C02. It is interesting to note that the degradation of the 12 wt % DEA in this run is much slower than it would have been if the THEED were not present. It appears that THEED, like BHEP, is able to reduce the availability of C02 and.HC03 for DEA attack. Figure 10.11 shows plots of the THEED concentration versus time for the two runs. The curves are initially linear and indicate a first order reaction of THEED forming BHEP. The asymptotic behaviour exhibited by the THEED concentration in run 69 is probably due to the fact that additional THEED is formed from the DEA degradation. The effect of CO, causes the k to increase about five-fold. Table 10.2 shows 1 THEED the values of k obtained from Fig. 10.11. It is interesting to THEED note that the value of k - under 4137 kPa (600 psi) CO. at 205°C is THEED 2 very close to that obtained from the kinetic model of the degradation of DEA which will be developed in chapter 12. 1 I 1 1 1 COMPOUND 0 1 2 3 4 5 6 7 TIME (hr) Figure 10.10 Concentration of DF,A, HEOD, THEED, and BHEP as a function of time (reactants—DEA and THEED, 4137 kPa CO,, 205°C) Figure 10.11 log[THEED] ns a function of time (reactants —DEA and TI IKED, 41.37 kPa C02 or N2 , 205°C) 142 Table 10.2 kTHEED for reactions under C02 and N2 Run Pressure kTHEED _i kTHEED m°del hr"1 No. Gas psi kPa hr 1 68 C02 600 4137 0.25 0.26 69 N2 600 4137 1 0.057 -10.5 Experiments to study the effect of impurities in the DEA feed The main impurities in the DEA feed are MEA and TEA. Four runs were carried out to determine whether these impurities degrade and whether they react with DEA to form any other degradation compounds. Table 10.3 summarizes the conditions of the runs performed. Table 10.3 Experimental conditions of runs performed to determine the effects of DEA feed impurities Concentration Run Wt % Temperature C02 Pressure No. MEA TEA DEA °C psi kPa 70 30 - - 205 600 4137 71 - 30 - 205 600 4137 72 10 - 20 205 600 4137 73 - 10 20 205 600 4137 The degradation of MEA produced mainly HEI with small amounts of 0ZD and HEED. It was also noticed that samples of degraded MEA solu tions which had been stored at room temperature for several months, smelt of ammonia. TEA was unaffected by C02 and no degradation compounds were detected. When DEA was degraded in the presence of MEA (run 72) several 143 new peaks were detected on the chromatogram. Figure 10.12 is a typical chromatogram from run 72, with all the major peaks labelled. The major degradation compounds produced as a result of the reaction between DEA and MEA under 4137 kPa (600 psi) C02 were found to be HEP, BHEED, and BHEI. Run 73, where DEA was degraded in the presence of TEA, produced little change from that using pure DEA. However a small peak was detected with a retention time of 20.6 minutes. This was identified to be TEHEED. The mechanism for the production of these new degradation compounds will be discussed in chapter 11. 144 DEA BHEI MEA BHEP THEED BHEED HEP WVrr 1 HEOD 1 HEI Figure 10.12 Typical chromatogram of a degraded solution of DEA and MEA CHAPTER 11 DEVELOPMENT OF A MECHANISM FOR DEA DEGRADATION In this chapter, the results of the experiments will be discussed and an attempt will be made to explain the observed phenomena. The overall purpose of this chapter is to develop a model for the degradation of DEA by C02. Reaction mechanisms are proposed which may explain the production of the compounds detected in the degraded DEA solutions. Certain reaction steps are, however, not fully confirmed since, in some cases, there was no way of testing their validity within the scope of this work. Thus certain aspects of the mechanisms remain proposals for explaining experimental observations. The chapter is split up into five sections. The first section deals with the formation and reactions of HEOD, THEED, and BHEP. In the second, the concept of three degradation routes (i.e., ionic, molecular, and thermal) is developed. The third section contains an explanation of observations which are at variance with the concepts proposed in the first two sections. The fourth section provides an explanation of the production of minor degradation compounds. In the fifth section the conclusions of this chapter are summarized and an overall model of DEA degradation is presented. 145 146 11.1 Formation and reactions of the major degradation compounds 11.1.1 Formation of HEOD. Referring to chapter 2 the following set of equilibria were established between C02 and DEA. Aqueous conditions: R2 NH + C02 R2 NH+COO" R2NC00 + H -t- [11.1] Non-aqueous conditions or high DEA concentrations: 2 R2NH + C02 R2NH COO + R2 NH R2NCOO H2+NR2 [11.2] Also DEA is able to react with the bicarbonate ion in the following manner:-R2NH2 + + HC03" R2NH + H+ ~ H R2NH2+HC03' R2NH2 + [11.3] R2NC00 + H" + H20 [11.4] In each case the amine carbamate ion is formed, R2NC00 linked either with H or H30 or R2NH2 By internal dehydration of the carba mate, HEOD can be produced:-0 R-N-C-0 . I \": C2H4- 0TH H+< -» R-N CH2 0 II C 0 + H20 CH, [11.5] DEA carbamate or, at high DEA concentrations:-0 " ! - +/ R-N-C7O ... H, HEOD 0 II C I \" C2H4-0-j21_ _ / NR, H R-N I CH, 0 + R2NH + H20 CH„ [11.6] HEOD DEA 147 11.1.2 Behaviour of HEOD under reaction conditions 11.1.2.1 Proof that BHEP is not produced directly from HEOD. 34 Polderman and Steele proposed that two molecules of HEOD react to form BHEP. 0 C x \ ch* -CH* R-N 0 CH2 - CH2 X \ j \ + f | • R-N N-R + 2C02 [11.7] CH2 CH2 0 N-R \ / \/ > CH2 - CH2 '2 c 2 HEOD BHEP This reaction seems unlikely since four bonds must be broken. Also the results of the experimental runs do not support this route. From the stochiometric equation of this route:-2 HEOD BHEP [11.8] the rate of production of BHEP then becomes d[^HEP] = k[HEOD]2 [11.9] dt Figure 11.1 shows the graphical relationship for plots of concentra tion versus time for HEOD and BHEP based on Eq. 11.9. For example, Eq. 11.9 states that if the concentration of HEOD is constant then the concentration of BHEP should increase linearly. In none of the experiments was this form of relationship observed. For example in run 65 (see Fig. 11.2) the concentration of HEOD falls sharply, whereas the concentration of BHEP rises with increasing slope (i.e., the rate of production of BHEP increases). Other examples can be found with the experiments conducted at 140°C 148 HEOD BHEP . A u z O u I • TIME Figure 11.1 Qualitative plots of concentration versus time, showing the possible relationships between HEOD and BHEP 149 TIME Figure 11.2 Sketch of the concentration of HEOD and BHEP as a function of time for run 65 150 and 150°C (see runs 10 and 8). Here the concentration of HEOD reaches a maximum then falls, whereas the BHEP concentration rises with increasing rate (see Fig. 11.3). Similarly, the runs using KHC03 instead of C02 (i.e., the ionic runs section 9.2) provide convincing proof that BHEP is not produced from HEOD. Again, in these runs the concentration of HEOD rises to a maximum and then falls, whereas the concentration of BHEP increases with increasing rate. Also in these runs the production of HEOD is much lower when using KHC03 instead of C02, but the production of BHEP is unaffected. What appears likely is that BHEP is being produced from THEED rather than HEOD. Finally when the long term run (see section 10.1, Fig. 10.2) is studied, it can be seen that the concentration of BHEP levels off. The only possible ways for BHEP to level off are either that BHEP comes to equilibrium with an intermediate (which has been shown not to take place, see section 10.2), or that the concentration of the intermediate falls to zero. The concentration of HEOD after a run time of 50 hr. is about -A 1.15 x 10 moles/cc and therefore unlikely to be the intermediate whereas the concentration of THEED has dropped virtually to zero. 11.1.2.2 Equilibrium between HEOD and DEA carbamate. Blanc 36 et al. suggested that HEOD is attacked by water which breaks the ring to form carbamic acid. The acid then reacts with a molecule of DEA to form THEED. 0 II C / \ R-N 0 + H.O • R.NCOOH [11.10] II CH2 - CH2 HEOD Carbamic acid TIME Figure 11.3 Sketch of the concentration of HEOD and as a function of time for runs 8 and 10 BHEP 152 10 i i II i H ill ., / R2-NrC-0;-H + jlO^-CjHu " N • R2N-C2H4-NR2 + H20 + C02 [11.11] Carbamic acid DEA THEED The problem with this proposal is that the carbamic acid is highly unstable and exists in aqueous solutions only as the carbamate ion R2NCOO . Since HEOD is formed from the carbamate ion, it seems unlikely that the next stage of degradation is for HEOD to revert back again to the carba mate ion. This can be seen in Equation 11.12 below. - + - + DFA DEA + C02 • R2NC00 H • HEOD • R2NGOO H THEED [11.12] What actually appears to be happening is that HEOD is in equilibrium with the carbamate ion and is not an intermediate in the production of THEED or BHEP. See Equation 11.13 below: DEA + CO, • R2NC00~H+ • THEED [11.13] U HEOD When there is a limited amount of water (i.e., at high DEA concen trations) , it may be possible for HEOD to become an intermediate according to the following equation. 2 DEA + C02 • R2NC00_H2+NR2 • HEOD + H20 + DEA [11.14] THEED * R2NCOOH + DEA « In the absence of water the carbamate can link with a molecule of DEA. Carbamic acid may then be formed from HEOD and react with DEA to form THEED. However, in the normal situation (i.e., DEA concentration < 40 wt %), HEOD is unlikely to be an important intermediate in the forma tion of either THEED or BHEP. 153 11.1.2.3 Proof that THEED is not produced directly from HEOD. Referring to run 65 where HEOD and DEA are reacted in the absence of C02, it is observed that the concentration of HEOD falls sharply then levels off (Fig. 10.6). If HEOD were just an intermediate its concen tration would tend to fall to zero. If it is assumed that THEED is produced from HEOD, then the kinetic relationship follows from the reaction:-k kfe 2 HEOD THEED • products [11.15] d[T^ED] = ka[HE0D]J - k^ [THEED] [11.16] The product k [HEOD]2 becomes constant when the concentration of HEOD becomes constant. When k^[THEED] equals ka[HE0D]2, the concentration of THEED must level off since —-—3 - becomes zero. This is clearly-not d t the case as can be seen from Fig. 11.4, which shows the general plot of THEED and HEOD concentrations versus time for the temperature experi ments (runs 1 to 12). 11.1.3 Proposed model for the production and reactions of HEOD. It has been shown that BHEP and THEED are not directly produced from HEOD. However, run 62 indicated that HEOD can react to form DEA, BHEP, and THEED. These observations may be explained by referring to the following model in which HEOD is in equilibrium with the DEA carbamate ion and this is, in turn, in equilibrium with DEA:-DEA + C02 ^ R2NC00_H+ • THEED • products [11.17] HEOD According to this scheme, heating HEOD causes it to first break down to the carbamate which, in turn, either reverts to DEA or reacts with itself or DEA to produce THEED. THEED can then in turn produce BHEP. Using this scheme is it possible to explain the behaviour of HEOD 154 TIME Figure 11.4 Sketch of the concentration of HEOD and THEED as a function of time for runs 1 to 12 155 under various operating conditions? An equilibrium is first established (very quickly) between DEA, C02, and R2NC00 H+. Next, a second equili brium is established between R2NC00 H+ and HEOD at a much slower rate. At the same time R2NC00 H+ is slowly removed by further degradation. Therefore, the concentration of HEOD should initially rise rapidly until an equilibrium level is established with R2NCOO and then start to fall as the level of R2NCOO falls. In some cases this was observed experi mentally. However, at high temperatures the concentration of HEOD was observed to rise rapidly and then level off or in some cases fall very slowly. Thus the above scheme is too simplistic and needs refinement. First consider the equilibrium reactions between DEA, C02, and water. All of the equilibria are established rapidly in comparison with the degrada tion reactions:-C02 +H20^ H+ + HC03~ [11.18] R2NH + H20 ^ R2NH2+ + 0H~ [11.19R2NH + H+ ^± R2NH2+ [11.20] C02 + 0H~ ^ HC03~ [11.21R2NH + HC03~ ^ R2NC00~ + H20 [11.22] R2NH + C02 ^ZT R2NH+C00~ [11.23R2NH+C00~ • R2NC00~ + H+ [11.24] Since the deprotonation of the zwitteron (Eqn. 11.24) is practically instantaneous^ the formation of the carbamate from C02 and DEA can be considered irreversible (Eqns. 11.23 and 11.24). The bicarbonate and the protonated DEA are able to form the amine bicarbonate which can establi a further equilibrium reaction with the carbamate.(Eq. 11.26). 156 R2NH2+ + HC03_ ^ [R2NH2 + ] [HCO3-] [11.25] [R2NH2 + ] [HCO3-] Z=± R2NC00_ + H+ + H20 [11.26] The last reaction (Eqn. 11.26) is considered to be much slower than the other reactions. This is based on the fact that in the ionic reactions (sect. 9.2), DEA degrades at a slower rate than normal. (This is discussed in sect. 11.2.1.) All these equations (Eqns. 11.18-11.26) can be simplified to the following relationship. R2NH+CO2+H20 ^£ R2NCOO_ + H+ + H20 S— R2NH2+ + HC03_ [11.27] It needs to be remembered that the solubility of C02 for most high temperature runs lies between 0.2-0.3 g C02/g DEA or about 0.45-0.7 moles C02/moles DEA. Therefore, DEA will initially be in excess for the high temperature runs. The reaction mixture will consist essentially of DEA either in the free or protonated form and C02 is tied up either as the bicarbonate or carbamate ion. Degradation appears to begin with the carbamate ion dehydrating to form HEOD and slowly an equilibrium is set up between HEOD and R2NC00 . If no additional degradation were to take place, then the HEOD concentra tion would level off. However, R2NC00 further reacts slowly to form THEED. Since the concentration of R2NC00~ falls it would be expected that HEOD would also decrease. This does not occur and thus some mechan ism must be keeping the concentration of R2NCOO constant. Since it . is assumed that all the C02 is tied up either as HC03 or R2NC00 , the extra R2NCOO is not being produced from the reaction between C02 and DEA. The formation of THEED, however, produces bicarbonate ions. These ions will then upset the right hand side of the equilibria, Eqn. 11.27, causing more R2NC00 to be produced from the HC03 reacting with the excess DEA. This process restores the level of R2NC00~ to its original 157 value and keeps the HEOD concentration constant. What actually happens is that the excess DEA is slowly converted to THEED via the formation of R2NC00 from R2NH2+ and HC03_ (Eqn. 11.26). When no more free DEA is available, the concentration of R2NC00 starts to fall and so will HEOD. Figure 11.5 summarizes the proposed mechanism. HEOD + 2 H20 C0.-K)H_ or H,0 R.NH+H+ or H20 1 _ i R2NH+C02+H20 —• R2NCOO~+H++H20 . HC03 + R2NH2 + products * THEED + HCO excess R2NH + H Figure 11.5 Schematic diagram of degradation of DEA. It must be remembered that the chromatographic analysis is unable to differentiate between R2NH, R2NH2+, and. R2NCOO~. This causes diffi culties in confirming the above mechanism and therefore it must remain for the present only a theory based on inference. 11.1.4 Effect of temperature on the production of HEOD. As pointed out in section 8.1.1.1, the maximum concentration of HEOD falls with increasing reaction temperature. There are possibly two reasons for this. The first is simply that the solubility and hence concentration of C02 in the reaction mixture falls with increasing temperature. There fore the levels of R2NCOO and hence HEOD also fall. The second reason is more subtle and may be due to the fact that the reaction forming THEED is more temperature sensitive than the reaction forming HEOD. Thus at high temperatures the formation of THEED from R2NC00 is increasingly favoured. This was confirmed by the kinetic model developed in chapter 12 (see Figs. 12.4 and 12.7). 158 11.1.5 Reaction of HEOD and DEA under N2. Using the proposed model it is possible to explain the results of run 65 (see Figs. 10.5 to 10.8). . Initially the concentration of HEOD will fall rapidly as the carbamate is formed. However, as the equilibrium is established the concentra tion of HEOD levels off. At the same time equilibria are established between R2NC00~, R2NH2+, and HC03~ (see Eqn. 11.26). Then R2NC00~ slowly reacts to form THEED and HC03 . Thus a cycle is set up where the DEA feed is able to combine with the HC03 produced by the formation of THEED, to form more R2NC00 . Therefore DEA is slowly consumed forming THEED with the concentrations of HEOD, R2NC00 , and HC03 all remaining relatively constant. 11.1.6 Formation of THEED. It seems unlikely that THEED is produced 36 from HEOD as proposed by Blanc et al. What is suggested here is that the carbamate ion reacts either with itself or a molecule of DEA to form THEED in the following manner. R iQ N7C-0 . . .H +H0;-C2Hi,-N R ^C-0 ...H+ R R \ / N-CjHu-N / \ - + R C-0 ...H + H++HC0, DEA carbamate 0 0 DEA carbamate THEED carbamate [11.28] .-CjHu-N \ R N-C2H^-N + H++HC0, R H R H [11.29] DEA carbamate DEA THEED In reaction 11.28 a carbamate ion is formed similar to the DEA carbamate ion. This THEED carbamate ion may then revert to THEED or 159 4 react further to form BHEP (discussed in section 11.1.7). In the absence of water or at high DEA concentrations it appears that HEOD may act as an intermediate in the formation of THEED as proposed 36 by Blanc et al. (see Eqns. 11.10 and 11.11). THEED can also be produced directly from DEA by the dehydration of two molecules of DEA. This reaction can be considered the 'thermal degradation' of DEA (see section 9.4). R RRR \ , , / \ / N r H + HO r C-Hu-N • N - C-H^-N + H,0 [11.30] / ' ' \ / \ R H R H DEA DEA THEED A third possible route, which can also be considered a thermal degradation route (see section 4.2) is where DEA loses a molecule of water to form an imine (HEM). The imine can then react with another molecule of DEA producing THEED. H CH, 2 R-N CH2 - CH2 •}• OH CH2 DEA HEM + H20 [11.31] CH2 —_ R R R R-N X ^ / \ / + H f N • N - CjH^-N [11.32] \ \ / \ CH, R H R HEM DEA THEED 160 11.1.7 Formation of BHEP. After studying the various plots of concentration versus time, it became evident that BHEP was not produced 34 from HEOD as suggested by Polderman and Steele . What is proposed here is that THEED dehydrates to form BHEP. R N - C2HK-N R R-N C2H, / \ N - R + H20 [11.33] C2H4 H CjH^OH THEED BHEP It was observed from the experiments (see sect. 10.4) that the rate of this reaction was increased considerably by the presence of C02 and HC03 . It seems likely that C02 increases the rate of conversion of THEED to BHEP in a similar manner to the degradation of DEA, i.e., via the formation of a carbamate. Thus C02 and HC03 not only catalyse the degradation of DEA to HEOD and THEED, but also the degradation of THEED to BHEP. R HOfC.H,, N-CjH^-N X R c-o" II 0 R-N .H C2H, C2H4 N-R +H +HC0, [11.34] THEED carbamate BHEP 11.2 Discussion of the degradation routes 11.2.1 Ionic route. The runs using KHC03 yielded similar degrada tion products to those formed in the normal C02 run. This indicates that HC03 aids the degradation of DEA in a similar way to C02. How ever, the rate of degradation due to HC03 is considerably lower and only small amounts of HEOD are produced. The reason for the lower rate of reaction is the fact that the amine salt (i.e., R2NH2 HC03 ) 161 must break down to form the amine carbamate. IH H-0! 0 /" \ II -R2N C = 0 • R2N-C-0 ... H + H20 [11.35] H • • • • 0 Under normal conditions the carbamate can be produced quickly by free C02 reacting with DEA. The proposed ionic route may be summarized by the following equations. R2NH + H20 ^ R2NH2+ + H0~ [11.36] 0H~ + C02 ^ HC03~ [11.37R2NH2+ + HC03~ [R2NH2 ] + [HC03 ]~ ;z± R2NCOO~H++H20 [11.38] R2NCOO~H+ ^ HEOD + H20 [11.39R2NC00~H++R2NH • R2 N-C2 H^-NRH + HC03~ + H+ [11.40] R2N-C2Hu-NRH • R-N-(C2Hu)2-N-R + H2 0 [11.41The low concentration of HEOD produced in the runs using KHC03 can be explained simply by the fact that the presence of KHC03 tends to increase the solution alkalinity as opposed to C02 which decreases the solution alkalinity. Under alkaline conditions HEOD becomes more unstable (see sect. 9.1). Therefore HEOD will exist at lower concentra tions in alkaline solutions, i.e., there is a shift in the equilibrium from HEOD towards the carbamate. Referring to Table 9.3 in section 9.2, it was noticed that the initial rate constants (k^^) for the ionic runs were virtually identical to those where solutions of 10 wt % DEA and lower were degraded with C02. From this it can be concluded that, at low concentrations of DEA the favoured degradation route is the ionic route. It must be remembered that the proportion of R2NH2+ to the total DEA concentration is at its maximum at low DEA concentrations. 162 11.2.2 Molecular route. Before discussing this route a distinc tion must be made between the so called "molecular runs" and the "molecular route." The molecular runs involved degrading DEA with C02 in the absence of water. The rate of degradation was slower than normal due to the fact that water was unable to aid the degradation. The "molecular route" denotes the route where DEA reacts directly with C02 in aqueous solutions to form the carbamate ion. This is comparable to the ionic route where the carbamate is produced from the ions R2NH2 + and HC03 . 11.2.2.1 Molecular runs. It was observed that DEA could degrade in the absence of water to produce HEOD, THEED, and BHEP. The proposed route is as follows:-R2NH + C02 ^ R2NH+C00_ [11.42] R2 NH + R2NH+C00~ • R2NC00~H2+NR2 [11.43R2NC00~H2+NR2 HEOD + R2 NH + H2 0 [11.44] HEOD + H20 • R2NCOOH [11.45R2NCOOH + R2NH • R2 N-C2 Hi*-NRH + H2 0 + C02 [11.46] R2N-C2H„-NRH • R-N-(C2H J2-N-R + H2 0 [11.47The overall reaction rate is much lower than that observed under normal conditions because more steps are involved and reaction 11.45 requires water to attack the HEOD ring structure. Water is present in the DEA feed only as a trace impurity, but, with the water produced in reaction 11.44, there is sufficient water to begin the degradation. As the degrada tion proceeds, more water is formed. The following equations summarize the molecular route under normal reaction conditions where water is present in excess. 163 RjNH + C02^=T R2NH+COO~ [11.48] R2NH+COO~ + H20 ^ R2NCOO~H+ + H20 [11.49R2NCOO~H+^ HEOD + H20 R2NCOO"H+ + R2NH • R2N-C2H4-NRH + HC03~ + H+ [11.50] R2N-C2H„-NRH y R-N-(C2Hu)2-N-R + H20 [11.5111.2.3 Thermal route. The third route for the degradation of DEA consists of DEA reacting with itself to produce THEED. This implies that the reaction should be second order with respect to DEA. However, the experimental results indicate that a first order reaction (see sect. 9.4) is taking place. Thus DEA may first be degrading to an intermediate, which then reacts with DEA to give THEED. A possible intermediate could be HEM as mentioned in sect. 11.1.6. The proposed thermal degradation route becomes:-CH, k a R2NH • R-N_ + H20 [11.52] . CH2 R2NH + R-N. CH2 k b R2N-C2H4-NRH [11.53] CH2 R2N-C2HU-NRH • R-N- (C2 Hi,) 2 -N-R + H20 [11.54] The rate of DEA degradation becomes:-- dl51Al = + kb[HEM])[DEA] [11.55] The chromatographic analysis did not positively detect HEM in the degradation mixture. If HEM was being produced it may be assumed that its concentration is very small and relatively constant. Thus the forma tion of HEM becomes rate controlling and the degradation of DEA becomes a pseudo first order reaction, i.e., 164 _ d[D|Ai = kt[DEA] [n>56] 11.3 Discussion of anomalous experimental observations 11.3.1 The relationship between initial kpEA and DEA concentration. Figure 11.6 is a typical sketch of initial k^^ versus DEA concentration. Three distinct regions are observed: Region I (0-10 wt % DEA) In this region the main degradation route appears to be the ionic route (see sections 9.2 and 11.2.1). Region II (10-30 wt % DEA) As the concentration increases, the proportion of DEA existing as R2NH2+ falls and the degradation route becomes a combination of the ionic and molecular routes with the molecular route gaining dominance. The overall k„. therefore becomes the sum of the k values for the two DEA parallel degradation reactions. kn„. = (k„4). . + (kn„.) . . [11.57] DEA DEA ionic DEA molecular Since the thermal route is so much slower than either the ionic or the molecular route its contribution to degradation should be negligible. Region III (30-100 wt % DEA) As the concentration of DEA continues to rise the concentration of water falls. Thus the reaction becomes limited by water and the rate decreases until, at 100%, the degradation becomes that proposed for the molecular runs, i.e., Eqns. 11.42 to 11.47. 11.3.2 Arrhenius plot. Referring to the Arrhenius plot in Fig. 8.5, it was observed that, at high temperatures, the data tended to deviate from the linear form and the measured rate constants (k__.) became much DEA smaller than the predicted ones. One reason for this deviation could 165 166 be simply that, at high temperatures, the C02 solubility decreases and the C02 concentration becomes limiting. Another possible reason is that the ionic route becomes increasingly important with rising tempera ture. This can be seen in the following sketch. (Fig. 11.7) of the Arr henius plots comparing ionic and molecular runs with standard runs. What seems to be occurring is the amount of R2NCOO produced by the molecular route falls with increasing temperature. This could be due to the zwitteron being converted back to DEA before it is. deprotonated to form the carbamate. Therefore the overall production of R2NCOO falls with increasing temperature and inhibits the overall degradation. In summary the formation of the carbamate ion becomes:-R2NH + C02 >• R2NCOO~H+ at low temps, (molecular route) R2NH + C02 R2NH+COO~ *• R2NCOO~H+ at high temps. (molecular route) R2NH + HC03~ ^ R2NCOO~H+ + H20 at all temps, (ionic route) 11.3.3 Log [DEA] versus time plots. At high temperatures it was observed that the semi-logarithmic plots of DEA versus time were linear only for a few hours and then began to deviate. This indicates that the initial pseudo first order degradation reaction of DEA became inhibited as the reaction progressed. This inhibition could be the result of the following. 1. At high temperatures the concentration of C02 is very close to the critical value of 0.2 g C02/g DEA. Any reduction in this level will cause the degradation rate to fall. 2. It has been shown that the presence of degradation compounds (especi ally BHEP) inhibits degradation at high temperatures by tying up some of the available C02. 3. As the degradation proceeds, the C02 is converted to HC03 via the 167 log kDEA l/T Figure 11.7 Sketch of the Arrhenius plots for ionic, molecular, and standard runs formation of the degradation products and R2NC00 . Therefore the production of additional R2NC00 must proceed through the ionic route which is slower than the overall rate. Thus the overall degradation-rate will tend to fall. 4. As the reaction proceeds the mixture becomes more acidic due to C02 and the reduction of DEA.(Although BHEP and THEED are alkaline, two moles of DEA are required to produce each mole of THEED and BHEP; hence the number of moles of alkaline species falls.)' Experiments have shown that reducing the pH reduces the degradation rate. 11.3.4 Explanation of the effect of pH. The experimental results show that increasing the pH increases the rate of degradation and reduces the production of HEOD. This can be explained as follows. • 1. Increasing the alkalinity tends to aid the deprotonation of the zwitteron formed by C02 reacting with DEA, whereas acid conditions tend to stabilize the zwitteron. R2NH + C02 Z=T R2NH+COO" + 0H_ • R2 NCOO~ + H2 0 [11.58] R2 NH + C02 R2NH+COO" + H+ • R2NH+COOH [11.59] Therefore increasing the pH tends to increase the level of carbamate and hence degradation. 2. Increasing the alkalinity increases the solubility of C02 via the bicarbonate formation. This increase in C02 availability will, under conditions where C02 is limiting, increase the degradation. 3. HEOD is attacked by hydroxyl ions, reacting to form the carbamate. Thus an increase in pH reduces the stability of HEOD and hence the concentration of HEOD. 169 11.4 The formation of minor degradation compounds Besides the production of the major degradation compounds HEOD, BHEP, and THEED, many other compounds have been detected by the chromato graphic analysis and mass spectrometry. These ''minor" degradation compounds are produced in low amounts and may be ignored when developing the kinetic model of the degradation. These compounds may result from the reaction of DEA with impurities in the feed and various thermal routes. 11.4.1 MEA degradation. Probably the major impurity in the DEA feed is MEA which undergoes degradation when subjected to C02 under high temperature and pressure. The degradation has been fairly well documented. 38 According to Polderman et al. MEA first degrades to OZD, probably via the formation of MEA carbamate (RNHCOO ). 0 II C / \ RNH, + CO, • H-N 0 + H,0 [11.60] I I CH2 CH2 MEA OZD OZD then reacts with another MEA molecule to form HEI. 0 0 II II C C X \ / \ H-N 0 + RNH2 • H-N N-R + H20 [11.61] . CH2 CH2 CH2 CH2 OZD MEA HEI 40 This route was later shown by Yazvikova et al. to consist of two stages. The OZD reacts with MEA to form the urea, BHEU. BHEU then undergoes dehydration to form HEI. 170 0 II c H-N CH2 OZD MEA H R \ H / -* -N-C-N / \ R H BHEU [11.62] HOJC.H 0 II c -»- H-N I CH, N-R + H20 CH, [11.63] BHEU HEI The lone pair of electrons from the nitrogen atom of the basic MEA molecule attack the electron deficient carboxyl atom in the OZD ring and open it up to produce BHEU. If the ring is opened by OH a carba mate is formed and it is possible that an equilibrium is set up between OZD and MEA carbamate. This equilibrium is equivalent to that proposed in this work between HEOD and DEA carbamate. BHEU may also be formed by MEA reacting directly with the MEA carbamate in the following manner. 0 R n / H N-CTO ... H ' N-R / ' " ' / R R \ s / N-C-N / \ H H + H20 [11.64] / / \ * H H MEA carbamate MEA BHEU The BHEU can then undergo dehydration to form HEI as in Eqn. 11.63. The next stage of the degradation is the hydrolysis of HEI to form HEED in the presence of OH ions. 0 II c / \ H-N N-R H R OH- \ y + H20 • ^N-C2H4-NN^ + HC03 [11.65] CH2 CH2 H H HEI HEED The final stage is the dehydration of HEED to form piperazine or P. H H C,Hu \ / / \ N-C,Hu-N • H-N N-H + H, 0 [11.66] ^ <L \  / H ^ C2Hu+OH CjHu HEED P HEED can also be formed directly by MEA reacting with MEA carbamate. (This is similar to DEA reacting with DEA carbamate to form THEED.) R ' ~ ~ 1 H R H \ ll . + ' / \ / - + N-C-0 ...H HO-C.H4-N • N-C.H4-N + HCO, +H / 1 ' \ / \ H H H H [11.67] MEA carbamate MEA HEED Finally HEED may be formed by the thermal degradation of MEA. ! H HO7C.Hu R H /" \ \ / R-N N-H • N-C,Hu-N + H.O [11.68] \ / / \ H H H H MEA MEA HEED The chromatographic analysis of a degraded MEA solution only indi cated the formation of OZD, HEI, HEED, and P. However, it failed to reveal BHEU. It was also not possible to analyse for urea and other substituted ureas. Either the ureas broke down in the chromatographic 172 column or they had no effect on the flame ionization detector. 11.4.2 Reaction between MEA and DEA. It is possible for MEA carbamate to react with DEA to form N,N-bis-(hydroxyethyl) ethylenediamine or BHEED. R. i0 l H R H / \ / i NNfc-0 .. .H+ + HO f C.Hi, - N • 'N-C,HI»-N' / ! \ / \ H R H R MEA carbamate DEA BHEED + HCO3- + H+ [11.69] BHEED can then dehydrate to form N-(hydroxyethyl) piperazine or HEP. R H C,Hu \ / / \ N-C?Hu-N • R-N N-H + H,0 [11.70] x' \ , \ / H ^- CjHufOH C2Hu BHEED HEP Another degradation route could proceed via the formation of a substituted urea. MEA carbamate may react with DEA to form NNN'-tris(hydroxyethyl) urea or THEU. R „ R R _ R \ s,-r--+--. / \ S / N-C7O ...H H-f-N >- N-C-N + H,0 [11.71] H R H R MEA carbamate DEA THEU The urea can then dehydrate to form NN-bis(hydroxyethyl) imidazoli-done or BHEI. 173 0 II R _ R C \ 8 / / \ N-C-N • R-N N-R + H.O [11.72] H — C2Hi,fOH CH2 CH2 THEU BHEI Hydroxyl ions can then catalyse the hydrolysis of BHEI to form NN-bis(hydroxyethyl) ethylenediamine or BHEED. 0 ll C R R / \ OH" \ / R-N N-R + H.O ^ N-C.H^-N + HC03 [11.73] I I / \ CH2 CH2 H H BHEI BHEED The compounds THEU, BHEI, BHEED, and HEP may all be produced by reactions similar to Eqns. 11.67 to 11-73, which are initiated by TEA carbamate reacting with MEA. The only difference being the direct forma tion of BHEED from DEA carbamate and MEA where another isomer of BHEED is formed. R .7, , H R H \ >% - • / \ / N-C-0 ... H HO-C.H^-N • N-C.H^-N + HCO, + H /' ' \ / \ R H R H [11.74] DEA carbamate MEA BHEED There is one further route leading to the formation of BHEED, i.e., the direct reaction between DEA and MEA. R R R R \ ----- / \ / N-C.H^fOH + H-J-N • N-C.H^-N + H.O [11.75] / \ / \ H H H H DEA MEA BHEED 174 BHEI, BHEED, and HEP were all detected in small amounts in degraded solutions of DEA. The formation of THEU is only suspected since this compound could not be detected with the present analytical technique. 11.4.3 Minor degradation compounds produced from DEA. Since 40 . BHEU can be formed from MEA and C02, it seems likely that N,N,N,N'-tetra(hydroxyethyl) urea or TEHEU may be formed from DEA and C02. The proposed route is:-R Q R R Q R \ II,-- +---./ \ II / N-C-0 ... H + H-N • N-C-N + H,0 [11.76] / ' \ / \ R R R R DEA carbamate DEA TEHEU TEHEU is unlikely to undergo further degradation since it has no labile hydrogen (i.e., free hydrogen) attached to the nitrogen atom. Again the production of this compound is only proposed since it could not be detected. One other degradation route between two molecules of DEA was pro-36 posed by Blanc et al. They suggested that two molecules of DEA react forming N,N'-bis(hydroxyethyl amino ethyl) ether or BHEAE with the loss of water. R R R R N-C,Hu-0H + H - 0 - C.HH-N —• N-C-H^-O-C.H^-N / ' \ / \ H H H H + H20 [11.77] DEA DEA BHEAE No standards for this compound could be obtained and therefore it could not be determined whether it was produced during the degradation of DEA. 11.4.4 The reaction between DEA and TEA. TEA itself does not undergo any measurable degradation. However, it is able to react with DEA to form NNNN-tetra(hydroxyethyl)ethylenediamine or TEHEED. Two routes are possible, one involving DEA carbamate the other pure DEA. R r„ R R R \ i ll i \ 'N-C-0 ... H + HO - C.Hu-N • N-C2H4-N + HC03 + H 'l i 2 \ / 2 \ 3 R R R R [11.78] DEA carbamate TEA TEHEED or R\ /* R\ /* NfH + HO 7 C H^-N • 1-C HrN + HO [11.79] / \ / \ R RRR DEA TEA TEHEED TEHEED was the heaviest compound detected in the analysis of de graded DEA solutions. 11.5 Summary The following section summarizes the principle conclusions of this chapter. 11.5.1 Conclusions of the degradation experiments. Table 11.1 gives the main degradation routes for the range of operating conditions studied. 176 Table 11.1 Principal DEA degradation routes under various conditions Temp °C DEA Concentration Wt % Total psi pressure kPa l Route Limiting compounds 90-175 0-10 600 4137 Ionic — 90-175 10-30 600 4137 Ionic+Molecular -90-175 30-100 600 4137 Mainly molecular H20 175-250 0-10 600 4137 Ionic+Thermal2 -175-250 10-30 600 4137 Ionic+Molecular+ Thermal co2 175-250 30-100 600 4137 Ionic+Molecular+ Thermal3 C02+H20 The routes are:-a) Ionic:- R2NH2++HC03~ • R2NC00~H+ • products b) Molecular:- R2NH+C02 • R2NC00~H+ • products c) Thermal:- R2NH • products 2 At high temperatures the thermal route will start to contribute to the degradation, although only to a small extent. 3At high temperatures and DEA concentrations (> 30 wt %), the ionic route contributes more to the degradation than at lower temperatures where the molecular route is responsible for most of the degradation. This is probably due to the reduction in the formation of the carbamate via the molecular route, because the zwitteron reverts back to DEA faster than it is deprotonated to form the carbamate. 177 11.5.2 Summary of the degradation reactions 11.5.2.1 Reactions involving DEA. 1) 2 DEA • THEED + H20 THEED • BHEP + H20 2) DEA • HEM + H20 HEM + DEA • THEED THEED • BHEP + H2 0 3) 2 DEA • BHEAE + H20 11.5.2.2 Reactions involving DEA and C02. 1) DEA + C02 —• DEA carbamate DEA carbamate • HEOD + H20 DEA carbamate + DEA • THEED + H2C03 2 DEA carbamate • THEED carbamate + H2 C03 THEED • BHEP + H2 0 THEED carbamate • BHEP + H2C03 2) DEA + C02 • DEA carbamate DEA carbamate + DEA »• TEHEU + H20 3) DEA + C02 • DEA carbamate DEA carbamate —»• HEOD + H20 HEOD + DEA • TEHEU 11.5.2.3 Reactions involving MEA. 1) 2 MEA • HEED + H20 HEED • P + H20 11.5.2.4 Reactions involving MEA and CQ2. 1) MEA + C02 • MEA carbamate MEA carbamate —>• OZD + H20 MEA carbamate + MEA —»• HEED + H2C03 HEED —• P + H20 178 11.5.2.5 Reactions involving MEA and DEA. 1) MEA + DEA • BHEED + H20 BHEED • HEP + H20 11.5.2.6 Reactions involving MEA, DEA, and C02. 1) MEA + C02 • MEA carbamate MEA carbamate + DEA —• BHEED + H2C03 BHEED • HEP + H20 or DEA + C02 —• DEA carbamate DEA carbamate + MEA • BHEED' + H2C03 BHEED' • HEP + H20 2) MEA + C02 • MEA carbamate MEA carbamate + DEA • TEHEU + H20 or DEA + C02 • DEA carbamate DEA carbamate + MEA • TEHEU + H20 TEHEU • BHEI + H20 BHE1 + 2 H20 • BHEED + H2C03 BHEED • HEP + H20 11.5.2.7 Reactions involving DEA and TEA. 1) DEA + TEA »- TEHEED + H20 11.5.2.8 Reactions involving DEA, TEA, and C02. 1) DEA + C02 • DEA carbamate DEA carbamate + TEA • TEHEED + H2C03 Note: Wherever the carbamate is used the ion is being referred to. Also H2C03 exists under operating conditions as HC03 and H+. 11.5.3 The degradation mechanism. Figure 11.8 shows the major reactions responsible for the degradation of DEA with C02. I R-N 0 + H,0 CO, + H,0 ^2 HCO, + H — HCO," CO, + OH (HEOD) CH, CH. (DEA) R,NCOO + H — or H,0~ or R,NH,H R,NH © I R.N - C - R.N I 0 (TEHEU) R,NH R,NH + H,0 Z=Z R,NH, + 0H~ I IR,NH, + ] {HCO,"| R,NH,+ + HCO," © R,NH or R.NCOO" R-N CH, RjNH • CH, (HEM) R.NH HRN-C2H,,-0-C2H»-NRH — — —» (BHEAE) R,N - C,HH - NRH + H + HCO, (THEED) / \ R-N N-R + H,0 \ / C2H» (BHEP) © © © major degradation route minor degradation route Molecular route Ionic route Thermal route Figure 11.8 Schematic diagram showing the possible routes for the degradation of DEA CHAPTER 12 KINETIC STUDIES 12.1 Development of a kinetic model The purpose of the model is to predict, quantitatively, the degrada tion of DEA and the production of"its degradation compounds. At times a model can be based on the stochiometric equations of the reaction. Unfortunately in the case of DEA the degradation reaction is extremely complex involving several equilibria, parallel and series reactions. Therefore it is necessary to simplify the scheme as presented in chapter 11, Fig. 11.8. From the experiments it was established that the initial degradation of DEA was pseudo first order with the Arrhenius relationship being obeyed up to about 175°C. Above 175°C the Arrhenius plot deviated from the linear form. A simple kinetic model based on initial k„, values could r DEA not predict this. Since, under industrial conditions it is unlikely that temperatures ever exceed 150°C, a kinetic model need only be applic able up to about 175°C. Above this, the predictions of the model may severely disagree with measurements. The model may also be simplified by removing the effect of C02. This can be done by assuming that the C02 concentration is constant or not limiting. This occurs when the C02 concentration is greater than 0.2 g C02/g DEA, i.e., at low temperatures and high total reaction 180 181 pressure. Thus the ranges of conditions covered by the following model Temp: 90-175°C DEA cone: 0-100 wt % C02 loading: > 0.2 g C02/g DEA Under these conditions the Arrhenius plots can be considered linear and the effect of C02 ignored. Finally the model has to deal with the effect of initial DEA con centration on k^^. Based on the assumption that DEA degradation was governed by a pseudo first order reaction, experiments showed that k^^ was not independent of the initial DEA concentration (see Fig. 8.15). The figure shows three distinct regions 0-10, 10-30, and 30-100 wt % DEA. The simplest way for the model to deal with this effect is to produce a series of Arrhenius plots similar to Fig. 8.16, which cover the DEA concentration range for each reaction of the kinetic model. The plots could be used to obtain the k value of each reaction at any given set of operating conditions. 12.1.2 Simplified degradation mechanism. The equilibrium reactions between C02 and DEA and the formation of R2NC00 is established within a matter of seconds. Therefore, these initial fast reactions may be ignored, when compared to the slow degradation reactions, since they are not rate controlling. For simplicity the model will consider R2NC00 as DEA. Also, since the ionic and molecular routes both result in identical degradation products they will be considered as one route. The thermal route can be ignored since it is much slower than the normal degradation. Thus the model can be simplified into the following set of equations. 182 k DEA ^! HEOD [12.1] k' k" DEA —»• THEED [12.2] k' ' ' THEED • BHEP [12.3] These equations still present a problem. It has been shown that the initial DEA degradation is governed by a pseudo first order reaction. Therefore the rate of DEA degradation should be represented by an equa tion of the form:-_ djDEAl = k[D£A] dt Unfortunately the Eqns. 12.1 to 12.3 do not show this, instead they suggest an equation of the form:-- diDEA] = (k+k")[DEA] - k'[HEOD] [12.5] dt To deal with this it was decided to make the production of HEOD an irreversible reaction. This could be justified since, at low temper atures the equilibrium between R2NCOO and HEOD is established slowly and the plots of concentration of HEOD versus time do not level off. Furthermore, the concentration of HEOD when compared to that of DEA is very much smaller and slight errors in the prediction of the concentration of HEOD should not affect the overall model. Thus, the degradation mechanism can be simplified as follows:-HEOD DEA [12.6] \ THEED • BHEP 183 It must be realized that this is not a stochiometric relationship but a kinetic relationship, which can be reduced to a model for predict ing the degradation of DEA. 12.2 Theory Using Eqn. 12.6 it is possible to write equations for the rate of change of the various compounds, i.e.:-1151*1 = - ki[DEA] - k2[DEA] [12.7] ^ODl . kl[DEA] [12.8] dt d[Tt?EED] = k 2 [ DEA ] - k3 [THEED] [12.9] dt d[BHEP] = k3[THEED] [12.10] dt As shown in Appendix D, these equations can be solved to give:-[DEA] = [DEA]Q e~(kl+kz)t [12.11] [HEOD] = [DEA] o rr-^- (1 - e-(ki+K2)T:) [12.12] t K1+K2 [THEED]t =[DEA]o k3_(^+k2) (e_(kl+k2)t - e_k3t) [12.13] _ rnT?Al k2 , k3 _-(k2+ki)t ki+k2 _-k3tN [BHEP] t - DEA] 0 (1 " k3_(kl+k2) e - k3_(kl+k2) * ) [12.14] Since in many cases the plots of THEED concentration versus time go through a maximum (Fig. 12.1), relationships can be derived for relating ki,k2 and k3 using t max and [THEED]max. Again details are given in Appendix D. £n(k3/k2+kx) M9 lt-i t max = —. j\——,—7— [12.15J ks-(k2+ki) i , iii k3-(ki+k2) [THEED]max _ k2 (k2+kis fl2 161 [DEA] ~ k2+ki v k3 184 imox TIME • Figure 12.1 Typical plot of [THEED] versus time In addition, kDEA = kl + k2 [12.17] 12.3 Calculation of the k values Using the experimental data of DEA, HEOD, THEED, and BHEP concen trations versus time and equations 12.7 to 12.17, the following methods were used to calculate ki, k2, and k3. 12.3.1 Method (A)—The plot of [THEED] vs. t goes through a maximum. 1. From the linear plots of log[DEA] versus time, an initial k^^ was calculated from the slope of the plot. 2. Using the results of HEOD concentration versus time and Eqn. 12.12, it was possible to calculate a value of kj. Several values of ki could be calculated for different times and concentrations and then averaged. Alternatively a plot of [HEOD] vs. J^-k (1 - e"(kl+k2)t) 185 could be made; the slope of this line is ki. 3. Using Eqn. 12.17, ki and kj^; k2 could be calculated. 4. Using the determined values of t max and [THEED]max, and Eqns. 12.15 and 12.16, it was possible to calculate k3. Trial and error were used to solve each equation and the value of k3 was optimized between both equations. 5. Using the values of ki, k2, and k3 with Eqns. 12.11 to 12.14 theor etical values of DEA, HEOD, THEED, and BHEP concentrations could be calculated. Figure 12.2 shows a typical plot of the model predictions (dashed lines) compared with the experimental results (points) using method (A) (see run 6) . 12.3.2 Method (B)--The plot of [THEED] vs. t does not go through  a maximum. Steps (l)-(3) are identical to those for method (A). 4. k3 may be calculated using the differential Eqn. 12.10. The rate of BHEP production can be determined from the slope of [BHEP] versus time at various times. By plotting (d[BHEP]/dt)t against [THEED] a straight line should be obtained whose slope will give k3. Alternatively k3 may be calculated directly from Eqn. 12.10 at various times and the results averaged. This method of calculating k3 may be used as a check on the value of k3 calculated by method (A). Figures 12.3 shows a typical plot of the model predictions compared with the experimental results for run 23. 3.0 2.5 2.0 1.5 1.0 0.5 EXPERIMENTAL COMPOUND O —DEA A— HEOD • —THEED • — BHEP THEORETICAL 5" A —••—9' • • 6 8 10 TIME (hr) 12 14 OO a Figure 12.2 Comparison between the experimental and theoretical values of DEA, HEOD, THEED, and BHEP concentrations as a function of time (run 6) 2 3 4 5 6 7 TIME (hr) Figure 12.3 Comparison between the experimental and theoretical values of DEA, HEOD, THEED, and BHEI' concentrations as a function of time (run 23) 188 12.4 Comparison of the experimental results with the predictions of  the model Tables of experimental results versus predicted results are presented in Appendix E. The model gave a very good prediction of the concentra tions of DEA, THEED, and BHEP for various reaction times (see Figs. 12.2 and 12.3). In the case of HEOD, the model tended to over-predict the concentration after a certain reaction time. This was to be expected since the model did not account for the reversible reaction between HEOD and DEA (or more correctly R2NC00_). 12.5 Application of the model Figures 12.4 to 12.6 show Arrhenius plots for ki, k2, and ka.- The values of the k's were determined from the experimental data using method (A) or (B). The plots for ki and k2 both conform to the three concentration regimes observed in the Arrhenius plot for k^^ (see Fig. 8.16). In general the lower curve covers concentrations ranging from 0 to 10 wt % DEA and the upper curve covers concentrations between 30 to 100 wt % DEA. For concentrations in the range of 10 to 30 wt % DEA there can be considered a series of curves between the two extremes. (See Fig. 12.7, which shows an example of finding the k value at 140°C for a 17 wt % DEA solution.) The Arrhenius plot for k3 (Fig. 12.6) is unaffected by DEA concen tration and tends to confirm the fact that BHEP is produced from THEED. Also this plot is a straight line even at high temperatures. It is interesting to note that if this plot is extrapolated to 205°C the k value obtained agrees very closely with the value of 0.25 hr 1 calculated from the results of run 69 (section 10.4) where THEED was degraded under C02 to BHEP. 189 191 192 To predict the degradation of a given DEA solution under a specified set of conditions, the method is illustrated by means of a numerical example. Let 17 wt % DEA be absorbing C02 at 14 0C at a total pressure of 4137 kPa (600 psi). 1. First determine the values of ki, k2, and k3 from the Figs. 12.4 to 12.6. For ki and k2 estimate the value of the 17 wt % Arrhenius plot which lies between the 10 wt % and 30 wt % to curves and read off the corresponding k value for the given temperature, i.e., 140°C (see Fig. 12.7). 2. Using these values of ki, k2, and k3 and Eqns. 12.11 to 12.14 the concentrations of DEA, HEOD, THEED, and BHEP can be calculated for any desired time. Figure 12.7 Sketch of Arrhenius plots for ki and initial DEA concentrations k2 at various CHAPTER 13 PURIFICATION OF DEGRADED DEA SOLUTIONS It is not possible to purify DEA by standard means such as distil lation since DEA and its degradation products have similar vapour pres sures. Also at atmospheric pressure, DEA degrades near its boiling point. In addition some of the degradation compounds distill over a range of boiling temperatures, which rules out the possibility of vacuum distillation. 13.1 Use of activated carbon Activated carbon filters have been used in several natural gas treating units to purify degraded DEA solutions. Usually a 5-10% slip stream of the DEA solvent is passed continually through an activated carbon filter. Although it has been claimed by some that the filters are very successful,^'^ their general effectiveness has yet to be proven. It appears that the activated carbon absorbs surface active compounds, which may be the cause of foaming, and may remove some dissolved heavy hydrocarbons and possibly some of the heat stable salts. There is however little evidence to date that the filters are able to remove any of the degradation compounds. In order to determine whether activated carbon can remove degrada tion compounds, samples from industrial filter units were tested and a series of experiments were conducted in the laboratory. 194 195 Figure 13.1 shows two typical chromatograms obtained from samples taken upstream and downstream of an activated carbon filter in a large gas plant. The similarity between the chromatograms clearly indicates that the filter was ineffective for removing degradation compounds. Samples of DEA solutions degraded in the laboratory were contacted with activated carbon for periods ranging from a few hours to a few weeks at room temperature as well as at 50°C. In none of these experiments was the activated carbon found to change significantly the concentration of the degradation compounds. An example of the results can be seen in Fig. 13.2. Although none of the major degradation compounds were removed, the degraded DEA solutions did change from a dark brown colour to a light yellow. It therefore seems that activated carbon is unable to remove HEOD, THEED, or BHEP from degraded DEA solutions. 13.2 Use of solvents Several experiments were conducted to find a solvent in which DEA was soluble and its major degradation compounds were not or vice versa. If a successful solvent is found then a possible purification method could be developed. Unfortunately the tests were generally unsuccess ful.Either DEA and its degradation compounds were all soluble or all were insoluble. The results are tabulated in Table 13.1. a) Sample taken upstream of filter Figure 13.1 Typical chromatograms of partially degraded DEA solutions taken upstream and downstream of an activated carbon filter located in a large gas plant 197 BHEP a) Sample before contact with activated carbon dBHEP r-1 b) Sample after contact with activated carbon Figure 13.2 Typical chromatograms of partially degraded DEA solutions under laboratory conditions; before and after contact with activated carbon 198 Table 13.1 Effect of various solvents on degraded DEA solutions Solvent Comments Acetonitrile CH3 CN DEA and degradation compounds insoluble. Water soluble. Furan C4H4O DEA and degradation compounds partially soluble. Pyridine C5H5N DEA and degradation compounds soluble. Chloroform CHCL3 DEA, HEOD, THEED soluble, BHEP partially soluble. Water soluble. Ethyl alcohol C2H5OH DEA and degradation compounds partially soluble. N-propyl alcohol C3H7OH DEA and degradation compounds partially soluble. 13.3 Removal of BHEP When most of the water is stripped off from a degraded DEA solution with a high BHEP -4 concentration (i.e., ^ > 5 10 moles/cc),BHEP starts to crystallize out at room temperature. It is useful to keep a small amount of water in the solution since it prevents DEA from solidifying (the melting point of DEA is 27-30°C). The crystals of BHEP can then be removed by vacuum filtration. The crystals usually have some viscous DEA adhering to them which can be washed off with propyl alcohol. 13.4 Removal of HEOD HEOD is easily attacked by OH ions to give R2NC00 . Therefore, a simple way to recover DEA from HEOD would be to add NaOH to the de graded DEA solution and apply heat to drive off C02 from the carbamate. However a further problem may result from the fact that addition of NaOH increases the C02 solubility, which may in turn increase the degradation 199 (see pH experiments, section 9.1). 13.5 Conclusion In conclusion, the present purification experiments met with little success. It is recommended that natural gas processing plants try to operate under conditions which minimize degradation, rather than try to purify heavily degraded solutions. CHAPTER 14 CONCLUSIONS AND RECOMMENDATIONS The overall reaction between C02 and DEA consists of two stages. The first stage is the rapid establishment of a complex equilibrium between DEA, C02 , HC03 , 0H~, R2NH2+, and R2NC00~. The second stage, called the degradation reaction, is very much slower and DEA is converted irreversibly to degradation compounds. The degradation reaction between DEA and C02 is complex and cannot be described by a simple stochiometric equation. The major degradation compounds are HEOD, THEED, and BHEP. Minor degradation compounds which were detected, are HEED, HEP, OZD, BHEED, HEI, BHEI, and TEHEED. Some of these compounds were produced by the reaction of DEA with impurities in the feed such as MEA and TEA. The degradation reaction proceeds via the formation of the DEA carbamate ion (R2NCOO ). The carbamate can be produced by two routes:-a) 'The molecular route', where C02 reacts directly with DEA to form the carbamate. b) 'The ionic route' where C02, in the form of HC03 , reacts with DEA in the form of R2NH2+ to give a salt. The salt can then degrade to give the carbamate. DEA can also degrade without the presence of C02 forming THEED and BHEP. This reaction is very much slower than the normal 200 201 degradation reaction involving C02. It appears that C02 acts as a catalyst, with C02 being neither produced nor consumed. 7. The carbamate is able to either form HEOD and set up an equilibrium or react with itself or a molecule of DEA to form THEED. 8. Finally THEED can degrade to form BHEP. This reaction also appears to be catalysed by C02. 9. The overall initial rate of DEA degradation is governed by a pseudo first order reaction. 10. The rate of DEA degradation is strongly affected by temperature. The Arrhenius plot confirms this. However at temperatures greater than 175°C, the plot deviates from the straight line behaviour. This was explained by a) C02 becoming limiting and b) the molecular route becoming modified at high temperatures. 11. The rate constant (^£A) is also strongly affected by initial DEA concentration. Three regions can be defined: a) 0-10 wt % DEA, where the main degradation route is ionic. b) 10-30 % DEA, where the It . sharply increases as the degradation DEA is a combination of molecular and ionic. c) 30-100 wt % DEA, where the main degradation route is molecular. The rate slowly falls as water becomes limiting. 12. The degradation rate is unaffected by C02 pressure provided the C02 concentration in the reaction mixture is greater than about 0.2 g C02/g DEA. 13. The degradation rate increases with increasing pH. This is probably due to a) increased C02 solubility and b) an increase in the concentration of the carbamate. 202 14. The degradation rate is inhibited by the presence of degradation products, especially BHEP, at high temperatures. This is probably due to C02 being tied up with the degradation products, which reduces the concentration of available C02 for DEA degradation tc less than 0.2 g C02/g DEA. 15. Using a simplified degradation model HEOD DEA THEED • BHEP it was possible to develop equations for predicting the degradation of DEA and production of degradation compounds. The model covered the ranges of 90-175°C, 0-100 wt % DEA for C02 concentrations greater than 0.2 g C02/g DEA. 16. Activated carbon was found to be incapable of removing any of the major degradation compounds. 17. BHEP can be partially removed by drying degraded DEA solutions and allowing BHEP to crystallize out. 18. DEA can be recovered from HEOD by adding NaOH to the degraded DEA solution and applying heat. 14.1 Practical implications of the present study a) The effect of temperature. The design and operation of DEA units must avoid the creation of elevated temperatures throughout the plant. The heat transfer surfaces of the stripper reboiler (especially when gas fired) are particularly prone to the formation of localized hot spots. To prevent such hot spots in operating plants, the DEA circulation through the stripper reboiler should be kept high and the steam or gas temperatures kept low. If, for some reason, the DEA circu lation should decrease, immediate action must be taken to reduce the steam pressure or fuel gas flow. There are two other sites where major degradation may take place. The first is within the heat exchanger that heats the rich amine stream with the lean amine stream. In some 13 cases the temperature of the rich amine stream may be as high as 125°C. The second site is at the base of the absorber. If the C02 content of the raw natural gas is high the temperature of the rich amine at the base of the absorber may rise to 110-120°C. In many DEA units only the bulk solution temperatures are measured It must be remembered that the skin temperatures of heat transfer surface can be very much higher, particularly during process upsets. Reliance on bulk temperatures is therefore inadequate. b) Effect of pressure. The partial pressures of C02 should be kept as low as possible in order to minimize DEA degradation. Although it is not usual to exercise control over the C02 content of the raw gas entering a plant, it may be possible to dilute the raw gas with some purified natural gas thus diluting the overall C02 content. This dilut would not only reduce the degradation rate, it would also reduce the heat of absorption when C02 is absorbed into DEA. This would help to keep the overall temperature in the absorber low. c) Effect of DEA concentration. Ideally the plants should try to operate with low DEA solution strengths (if possible well below 20 wt %). However, limitations are imposed by the desired plant capacity. Future design of gas treating plants should consider larger equipment for operation with dilute solutions of DEA and dilute raw gas feed. 204 Again a dilute solution of DEA would reduce local increases in temperature, due to the C02 absorption. However, studies would have to be made to determine the cost effectiveness of these measures. d) Effect of activated carbon filters. Although the activated carbon appears unable to remove the major degradation products, it is not recommended to remove the filters from existing plants. The filters may serve other useful functions such as removing surfactants which can cause foaming, heat stable salts which may cause corrosion and they can act as a means for removing fine particulates. e) Purification of degraded DEA solutions. Since purification of DEA solutions is very difficult it is recommended that rather than provide equipment to purify the solutions the plant should be built and operated in such a way so as to minimize degradation. f) Analytical technique. The chromatographic analytical technique developed in this study is ideally suited for plant use. The method is simple and relatively fast with no sample preparation required. Using this method, it is possible to monitor DEA streams regularly. If degrada tion occurs, it is easy to detect and appropriate action can quickly be taken to minimize the DEA losses. 14.2 Experimental recommendations a) Measurement of pH. DEA degradation appears to be affected by solution pH measured at room temperature. It would, therefore, be useful to measure pH at the high temperature and pressure of a typical degradation experiment. b) Measurement of DEA carbamate concentration. The proposed degra dation of DEA appears to proceed via the production of DEA carbamate. To further clarify the degradation mechanism it would be useful to determine 204a the concentration of the DEA carbamate. However, using the existing chroma tographic analytical technique it is impossible to detect the carbamate since it is unstable and reverts back to DEA. Silylation may stabilize the carbamate sufficiently to allow its concentration to be determined using chromatography. 205 Symbol A BHEAE BHEED BHEI BHEP BHEU BHG DEA DEEA E ED EO GC/MS HEED HEI HEM HEOD HEP k kDEA ki,k2,k3 MDEA MEA NOMENCLATURE Explanation and typical units Frequency factor in the Arrhenius Eq. 8.7 (hr Bis (hydroxyethylaminoethyl) ether N,N-Bis(hydroxyethyl) ethylenediamine N,N-Bis(hydroxyethyl) imidazolidone N,N-Bis(hydroxyethyl) piperazine N ,N-Bis(hydroxyethyl) urea N,N-Bis(hydroxyethyl) glycine Diethanolamine Diethyl ethanolamine Activation energy in the Arrhenius Eq. 8.7 (Cal/g mol) Ethylenediamine Ethylene oxide Gas chromatograph with mass spectrometer N-(hydroxyethyl) ethylenediamine N-(hydroxyethyl) imidazolidone N-(hydroxyethyl) ethylenimine 3-(hydroxyethyl)-2-oxazolidone N-(hydroxyethyl) piperazine Reaction rate constant Overall reaction rate constant for the degradation of DEA, (hr"1) Rate constants used in the kinetic model of the degradation of DEA, Eqns. 12.7-12.17, (hr_1) Me thy1d ie thano1amine Monoethanolamine 206 OZD Oxazolidone P Piperazine R- -C2H4OH T Absolute temperature (°K) t Time (hr) TEA Triethanolamine TEHEED N,N,N,N-Tetra(hydroxyethyl) ethylenediamine TEHEU N,N,N,N-Tetra(hydroxyethyl) urea THEED N,N,N-Tris(hydroxyethyl) ethylenediamine THEU N,N,N-Tris(hydroxyethyl) urea [ ] Denotes concentration (g mol/cc) [ ] Denotes concentration at time t (g mol/cc) 207 REFERENCES 1. Scheirman, W.L,, Hydrocarbon Processing, 53(7), 95 (1973). 2. Butwell, K. F., Kubek, D. J., and Sigmund, P. W., Union Carbide Publi cation "Primary vs Secondary Amines," (1979). 3. Butwell, K. F., Kubek, D. J., and Sigmund, P. W., Hydrocarbon Process ing, 62(3), 108 (1982). 4. Jones, V. W., and Perry, C. R., "Fundamentals of Gas Treating," Proc. Gas Cond. Conf., 23(E), 1 (1973). 5. Younger, A. H., "Processes for Sour Natural Gas Treating," Symp. on Sulphur recovery from hydrocarbon feedstocks, Vancouver, B.C., Sept. (1973). 6. Swaim, C. D., Hydrocarbon Processing, 49(3), 127 (1970). 7. Dankwerts, P. V., and Sharma, M. M., The Chemical Engineer, October CE244 (1966). 8. Butwell, K. F., and Perry, C. R, "Performance of Gas Purification Systems Utilizing DEA Solutions," Proc. Gas Cond. Conf., 25(B), 1 (1975). 9. Beck, J. E., "Diethanolamine—An Energy Conserver," Proc. Gas Cond. Conf., 25(A), 1 (1975). 10. Heisler, L., and Weiss, I. H., "Operating Experience at Aderklaa with Alkanolamine Gas Treating Plants for Sour Natural Gas Sweeten ing," Proc. Gas Cond. Conf., 25(H), 1 (1975). 11. Smith, R. F., and Younger, A. H., "Operating Experiences of Canadian Diethanolamine Plants," Proc. Gas Cond. Conf., 22(E), 1 (1972). 12. Smith, R. F., and Younger, A. H., Hydrocarbon Processing, 51(7), 98 (1972). 13. Katz, D. L., Cornell, D., Kobayashi, R., Poettman, F. H., Vary, J. A., Elenbaas, J. R., and Weinang, C. F., "Handbook of Natural Gas Engineering," McGraw-Hill Book Co., Inc., N.Y., 1959. 14. Kohl, A., and Riesenfeld, F., "Gas Purification," McGraw-Hill Book Co., Inc., N.Y., 1960. 208 15. Dow Chemical Company, "Gas Conditioning Fact Book," The Dow Chemical Co., Midland, Michigan, 1962. 16. Nonhebel, G., "Gas Purification Processes for Air Pollution Control," 2nd Edition, Newness-Butterworths, London, 1972. 17. Union Carbide, "Test Procedures for Gas Treating Chemicals," Union Carbide Canada Ltd., Calgary Lab., 1974. 18. Mahajani, V. V., and Dankwerts, P. V., Chem. Eng. Sci., 37(6), 94'3 (1982) . 19. Ratkovics, F., and Horvath, I., Hungarian Jnl. of Ind. Chem., 9, 281 (1981). 20. Alvarez-Fuster, C, Midoux, N. , Laurent, A., and Charpentier, J. C, Chem. Eng. Sci., 36(9), 1513 (1981). 21. Astarita, G., Savage, D. W., and Longo, J. M. , Chem. Eng. Sci., 36(4), 581 (1981). 22. Laddha, S. S., and Dankwerts, P. V., Chem. Eng. Sci., 36(4), 479 (1981) . 23. Dankwerts, P. V., Chem. Eng. Sci., 34, 443 (1979). 24. Hikita, H., Asai, S., Ishikawa, H., and Honda, M., Chem. Engng. Jnl., 13, 7 (1977). 25. Sada, E., Hidehiro, K., and Butt, M. A., The Canadian Jnl. of Chem. Engng., 54, 421 (1976) . 26. Coldrey, P. W., and Harris, I. J., The Canadian Jnl. of Chem. Engng., 54, 566 (1976). 27. Leder, F., Chem. Eng. Sci., 26, 1381 (1971). 28. Shrier, A. L., and Dankwerts, P. V., I and EC Fund., 8(3), 415 (1969). 29. Nunge, R. J., and Gill, W. N., A.I. Ch. E. Jnl., 9(4), 469 (1963). 30. Vidaurri, F. L. , and Kahre, L. C, Hydrocarbon Processing, 56(11), 333 (1977). 31. Sigmund, P. W., Butwell, K. F., and Wussler, A. J., Hydrocarbon Processing, 60(5), 118 (1981). 32. Blanc, C, Elgue, J., and Lallemand, F., Hydrocarbon Processing, 60(8), 111 (1981). 33. Jorgensen, E., Acta. Chem. Scand., 10, 747 (1956). 209 34. Polderman, L. D., and Steele, A.' B., Oil and Gas Jnl., 54(65), 206 (1956). 35. Hakka, L. E., Singh, K. P., Bata, G. L., Testart, A. C., and Andrejchyshyn, W. M., Gas Processing/Canada, 61(1), 32 (1968). 36. Blanc, C., Grail, M. , and Demarais, G., "The part played by Degrada tion Products in the Corrosion of Gas Sweetening Plants using DEA and MDEA," Proc. Gas Cdnd. Conf., 32, March (1982). 37. Gough, G. G., "The Analysis and Analytical Interpretation of Gas-Conditioning Diethanolamine Solutions," Proc. Natural Gas Process ing Assoc., Calgary (1969). 38. Polderman, L. D., Dillon, C. P., and Steele, A. B., Oil and Gas Jnl., 54(2) , 180 (1955). 39. Berlie, E. M., and Estep, J. W., Chem. Engng. Prog., 61(4), 82 (1965). 40. Yasvikova, N. V., Zelenskaya, L. G., and Balyasnikova, L. V., Zhurnal Prikladnoi Khimii, 48(3), 674 (1975). 41. Choy, E. T., Masters Thesis, Univ. of B.C. (1978). 42. Choy, E. T., and Meisen, A., "Degradation of DEA Treating Solutions," Proc. Canadian Natural Gas Processing Assoc., Calgary, Nov. (1977). 43. Kennard, M. L., and Meisen, A., Hydrocarbon Processing, 59(4), 103-(1980) . 44. Kennard, M.L., and Meisen, A., "Practical Aspects of DEA Degradation Studies," Proc. Canadian Gas Processors Assoc., Calgary, Sept. (1981) . 45. Meisen, A., and Kennard, M. L., Hydrocarbon Processing, 61(10), 105 (1982). 46. Henry, M. S., and Grennert, M., Petroleum Refiner, 34(6), 177 (1955). 47. Henry, M. S., and Grennert, M., California Natural Gasoline Assoc. Proc, 28, 64 (1955). 48. Brydia, L. E., and Persinger, H. E., Anal. Chem., 39(11), 1318 (1967). 49. Piekos, R. , Kobyiczyk, K., and Grzybowski, J., Anal. Chem., 47(7), 1157 (1975). 50. Saha, N. C, Jain, S. K., and Dua, R. K. , Chromatographia, 10(7), 368 (1977). 51. Van Wijk, R. , Jnl. of Chrom. Sci., 8^, 418 (1970). 210 52. Choy, E. T., and Meisen, A., Jnl. of Chromatography, 187, 145 (1980). 53. Zimmerman, H. K., and Cosmatos, A., Jnl. of Chromatography, 7, 77 (1962). 54. Savignac, A., Lattes, A., and Carles, J., Jnl. of Chromatography, 7, 261 (1962). 55. Sargent, R., and Rieman, W., Anal. Chim. Acta., 17, 408 (1959). 56. Lynes, A., Jnl. of Chromatography, 23, 316 (1966). 57. Gehrke, C. W., Leimer, K., Jnl. of Chromatography, 57, 219 (1971). 58. Di Corcia, A., and Samperi, R., Anal. Chem., 46(8), 977 (1974). 59. Applied Science Lab. Inc., "Use of Pennwalt 223 Amine Packing," Technical Bulletin No. 21, (1979). 60. S.G.E., "Analysis of Ethanolamine," S.G.E. Technical Publication on Glass Lined Tubing, (1978). 61. Elderfield, R. C, "Heterocyclic Compounds," John Wiley and Sons Inc., 5, 396. 62. "Kirk-Othmer Encyclopedia of Chemical Technology," 2nd Ed., John Wiley and Sons Inc., 10, 914. 63. Ishikawa, A., Tomomichi, B., Yamasaki, S., and Yasuo, S., Nippon Kagakie Kaishi, 8, 1592 (1974). 64. McKay, A. F., and Braun, R. 0., Jnl. of Org. Chem., 16, 1829 (1951). 65. Weinberger, A., "Heterocyclic compounds with 3 or 4 Member Rings," Vol. 19 in the series "The Chemistry of Heterocyclic Compounds," John Wiley and Sons Inc. 66. Matveev, I. S., and Potitun, N. N., Khim. Geteralskii Soedin, Akad. Nauk. Latv. SSR, 3, 465 (1965). 67. Wilson, A. C., U.S. Patent 2, 475, 068, July (1949). 68. "Kirk Othmer Encyclopedia of Chemical Technology," 2nd Ed., John Wiley and Sons Inc., 8, 528. 69. Levenspiel, 0., "Chemical Reaction Engineering," 2nd Ed., John Wiley and Sons Inc., 70 (1972). ». 70. Morrison, R. T., and Boyd, R. N., "Organic Chemistry," 3rd Ed., Allyn and Bacon, Inc., (1973). 71. Lee, J. I., Frederick, D. 0., and Mather, A. E., Jnl. of Chem. Eng. Data, 17(4), 465 (1972). 211 72. Lawson, J. D., and Garst, A. W., Jnl. of Chem. Eng. Data, 21(1), 20 (1976). 73. Millar, I. T., and Springall, H. D., "Sidgwick's Organic Chemistry of Nitrogen," 3rd Ed., Clarendon Press Oxford, (1966). 74. Budzikiewicz, H. , Djerassi, C, and Williams, D. H. , "Mass Spectro-scometry of Organic Compounds," Holden-Day Inc., (1967). 75. Beynon, J. H., Saunders, R. A., and Williams, A. E., "The Mass Spectra of Organic Molecules," Elsevier Publishing Co., (1968). 212 APPENDIX A Sources of Equipment and Chemicals a) Equipment Item Supplier Model Chromatograph Gas Chromatograph Syringe Septa Column Mass spectrometer  Autoclave Reactor Temperature controller Digital thermometer Chart recorder b) Chemicals Chemical Supplier Acentonitrile Mallinckrodt Ltd., Paris, Kentucky BHG BDH Biochemicals Ltd., Poole, England BHEED ICN Pharmaceuticals Inc., Plainview, N.Y. BHEI Frinton Laboratories, Vineland, N.J. BHEP Aldrich Chemical Co. , Milwaukee, Wis. co2 Union Carbide, Vancouver, B.C. Chloroethanol Aldrich Chemical Co., Milwaukee, Wis. Chloroform Caledon Laboratories Ltd., Georgetown, Ont. DEA Matheson Coleman and Bell, Norwood, Ohio DEE A Aldrich Chemical Co., Milwaukee, Wis. ED Aldrich Chemical Co., Milwaukee, Wis. EO Matheson of Canada Ltd., Whitby, Ont. Ethyl alcohol Mallinckrodt Chemical Works, St. Louis, Mo. Ethyl carbonate Eastman Kodak Company, Rochester, N.Y. Furan Aldrich Chemical Co., Milwaukee, Wis. HCL Allied Chemicals Canada Ltd., St. Claire, Quebec HEED Aldrich Chemical Co., Milwaukee, Wis. HEI Frinton Laboratories, Vineland, N.S. HEOD Synthecon Laboratories, Vancouver, B.C. HEM Aldrich Chemical Co., Milwaukee, Wis. HEP Aldrich Chemical Co., Mulwaukee, Wis. MEA Malinckrodt, St. Louis, Mo. Hewlett Packard, Alvondale, PA. 5830 Hamilton Co., Reno, Nev. 701 Supelco Inc., Bellefonte, Penn. Microsep F-174 Supelco Inc., Bellefonte, Penn." Tenax G.C. Hewlett Packard, Vancouver 5985 B Parr Instrument Co., 111. 4560 Parr Instrument Co., 111. 4831 EB Doric Instruments, Que. Series 400A, No. 410A Corning 840 Appendix A (cont.) Chemical Supplier MDEA NaOH OZD Potassium bicarbonate Potassium carbonate Pyridine N-propyl alcohol Silica gel TEA TEHEED Aldrich Chemical Co., Milwaukee, Wis. Fisher Scientific Co., Fairlawn, N.J. Mallinckrodt Chemical Works, St. Louis, Mo. Mallinckrodt Chemical Works, St. Louis, Mo. Mallinckrodt Inc., Paris, Kentucky Anachemia Chemicals Ltd., Toronto Aldrich Chemical Co., Milwaukee, Wis. ICN,K and K Laboratories Inc., Cleveland, Ohio 214 APPENDIX B Experimental results for the degradation of DEA by CO, Although certain other degradation compounds were detected in degraded solutions of DEA, only data on DEA, BHEP, HEOD, and THEED are recorded here. This was because the other minor degradation compounds and certain feed impurities, such as MEA, existed in very low concentrations and, therefore, they could be ignored when developing the kinetic model. TABLE B.1 RUN 1 : 30 WT% DEA, 250C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.00 _ _ 0.50 2.32 0.01 6 0.063 0. 180 1 .00 1.81 0.091 0.073 0.200 1 .67 1 .37 0.131 0.088 0.250 2.50 1 .04 0.323 0.075 0.220 3.50 0.72 0.391 0.085 0. 120 5.00 0.62 0.553 0.085 0.020 6.50 0.48 0.663 0.086 - . 7.83 0.38 0.695 0.091 -kDEA = 0.691 hr-1 TABLE B.2 RUN 2 : 30 WT% DEA, 225C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.00 _ 0.58 2.38 - 0. 125 0.210 1 .00 2.24 0.019 0. 140 0.460 1 .67 1 .76 0.080 0. 135 0.600 2.67 1 .44 0.195 0. 125 0.710 3.58 1 .05 0.311 0.135 0.700 5.00 0.82 0.412 0.125 0.441 6.00 0.68 0.459 0. 120 0.335 8.75 0.57 0.541 0.118 0.171 kDEA = 0.399 hr-1 TABLE B.3 RUN 3 : 30 WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.02 _ _ 1 .00 2.24 0.010 0.170 0.360 2.00 1 .63 0.051 0.200 0.640 3.00 1 .26 0.111 0.195 0.840 4.00 0.95 0. 170 0.203 0.820 5.00 0.78 0.245 0.203 0.700 6.00 0.71 0.332 0.208 0.650 7.00 0.65 0.396 0.208 0.576 8.00 0.57 0.423 0.208 0.500 kDEA = 0.291 hr-1 TABLE B.4 RUN 4 : 30 WT% DEA, 195C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.14 mm-1 .00 2.30 - 0. 188 0.280 2.00 1 .89 0.025 0.233 0.610 3.00 1 .47 0.071 0.250 0.870 4.00 1.21 0.104 0.228 0.870 5.00 0.91 0. 156 0.220 0.790 6.00 0.82 0.215 0.243 0.760 8.00 0.74 0.256 0.220 0.720 8.00 0.63 0.292 0.215 0.680 kDEA = 0.23 hr-1 a PJ > tr I oooooooooooo oooooooooooo OOO to ro ro w uiff»(»-'ro*>uia)owmo focno>-**»*»ao*»oo>^j*» ooooooooo III rowrjoooooo 00 N) O VD VD mm w-i^.OO>OOK)K)COCOai ooooooooooo cncncn*»*»*>>t»<&>cofo-> *»cn*»oococr\t\>-'cr\cncn ooocnoocnoooo oo-*oooooooo U) O U) CD 03 vl OMtk N) -• ovocn-'a^cnocncncoco ooooooooooo > S *0 IT" PJ a PJ > o o w z a: o PJ • TJ 3: 0 f PJ cn ac 0 PJ 0 o a » 0 1 1 CO -3 X PJ PJ O W C z -3 > dd tr1 PJ w to o O PJ > cn O CO Cu CT> O O T> CO O O to a PJ > CO tr 1 CO 00 Vj OMJ1 ,(» W W - O tr 31 OOOOOOOOO OOOOOOOOO f PJ 00 — -» — — to to CO a PJ COVO-«tO*»-J-«0>iO > 00 00—'OOOOOOOOOOO 0 0 OOOOOO w z X n _»_._. 0 0 0 PJ • — 00 co — \D — KD CO tO -J 3 0 0 0 0 cn cn O tr1 PJ in 00000000 X \ O PJ 0 co co co co co co to -> O -'OOOOO-J00 a X Ul-O-O-UlO —* O 1 1 co OOOOOOOO X vovovDvoco-ocnco PJ ocncototoo-Jo PJ toooooocno 0 TABLE B.7 RUN 7 : 30 WT% DEA, 162C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.00 3.000 _ _ 1 .00 2.803 - 0.105 0.090 2.00 2.620 - 0.200 0.200 3.00 2.450 0.0101 0.275 0.290 4.00 2.280 0.0150 0.320 0.388 5.00 2. 137 0.0201 0.364 0.460 6.00 2.000 0.0254 0.412 0.561 7.00 1.870 0.0350 0.420 0.653 8.00 1 .744 0.0420 0.425 0.731 kDEA = 0.0678 hr-1 TABLE B.8 RUN 8 : 30 WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.000 _ _ 5.00 2.600 0.0098 0.250 0.075 10.00 2.200 0.0210 0.350 0.150 15.00 1 .860 0.0380 0.470 0.400 20.00 1 .560 0.0425 0.510 0.652 25.00 1.310 0.0550 0.550 0.751 30.00 1.110 0.0640 0.570 0.900 40.00 0.840 0.0860 0.570 1 .075 50.00 0.710 0.1060 0.542 1 .225 60.00 0.580 0.1300 0.540 1 .200 71 .40 0.570 0.1500 0.541 1 .060 81.10 0.455 0.1860 0.540 0.910 95.00 0.420 0.2100 0.500 0.810 kDEA = 0.0316 hr-1 TABLE B.9 RUN 9 : 30WT% DEA, 145C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.200 — _ 3.25 3.000 - 0.215 -13.75 2.430 - 0.488 0.250 19.75 2. 160 0.0173 0.580 0.403 28.50 1 .740 0.0210 0.600 0.551 39.50 1 .430 0.0620 0.621 0.790 48.50 1 .360 0.0740 0.600 0.881 60.00 1 .050 0.1040 0.605 0.925 69.00 0.920 0.1240 0.580 0.944 kDEA = 0.0195 hr-1 TABLE B.10 RUN 10 : 30WT% DEA , 140C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.000 — _ _ 10.00 2.850 - 0.250 0. 120 20.00 2.540 0.0120 0.450 0.320 30.00 2.260 0.0210 0.575 0.504 40.00 2.000 0.0330 0.660 0.581 50.00 1 .754 0.0415 0.710 0.700 60.00 1 .605 0.0510 0.710 0.751 80.00 1 .270 0.0700 0.660 0.863 100.00 1 .010 0.1050 0.700 0.950 130.00 0.830 0.1550 0.700 0.910 154.80 0.670 0.1700 0.610 0.892 178.00 0.560 0.2060 0.580 0.810 201.00 0.530 0.2560 0.575 0.754 kDEA = 0.0115 hr-1 TABLE B.11 RUN 11 : 30WT% DEA, 120C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3. 150 _ _ 20.00 3.000 - 0. 150 -40.00 2.900 - • 0.300 0.056 60.00 2.800 - 0.380 0.089 80.00 2.650 - 0.525 0.100 100.00 2.500 0.0060 0.600 0. 140 120.00 2.340 0.0084 0.705 0.225 140.00 2.244 0.0100 0.751 0.240 160.00 2.131 0.0180 0.760 0.313 180.00 2.000 0.0250 0.770 0.345 200.00 1 .905 0.0310 0.780 0.375 kDEA = 0.0026 hr-1 TABLE B.12 RUN 12 : 30WT% DEA, 90C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.530 — — _ 100.00 3.500 0.050 -160.00 3.420 0.090 -300.00 3.455 0.155 -441.00 3.403 0.191 -511.00 3.361 0.216 0.014 631.00 3.310 0.250 0.048 700.00 3.280 0.290 0.060 kDEA = 0.000142 hr-1 TABLE B.13 RUN 13 : 100WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 10.00 — _ — 1 .08 8.00 0.069 0.255 3.200 2.67 6.41 0. 172 0.275 3.860 3.92 4.60 0.282 0.263 4.070 4.92 3.71 0.370 0.250 3.645 6.00 2.84 0.524 0.245 3.540 7.00 2.56 0.643 0.247 2.650 7.93 2.12 0.754 0.225 1 .958 kDEA = 0.195 hr-1 TABLE B.14 RUN 14 : 80WT% DEA , 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 7.80 - — -0.78 5.98 0.0224 0. 190 1 .525 2.00 4.62 0.0960 0.265 2.535 3.00 3.31 0.2220 0.275 3.070 4.00 2.20 0.3260 0.248 3.000 5.00 1 .93 0.4980 0.263 2.725 6.00 1 .48 0.6560 0.240 2.340 6.95 1 .34 0.8200 0.250 1 .980 8.58 1 .09 0.8700 0.235 1 .258 kDEA = 0.277 hr-1 TABLE B.15 RUN 15 : 60WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 6.18 — _ _ 1.00 4.18 0.0425 0.195 1 .500 1 .87 3.42 0.0800 0.225 1 .950 3.00 2.68 0.1710 0.240 2.030 4.00 1 .80 0.2560 0.225 1 .834 5.05 1 .36 0.3580 0.222 1 .560 6.08 1.15 0.4490 0.205 1 . 168 6.92 1 .06 0.5620 0.210 1 . 169 8.00 0.95 0.6370 0.200 0.954 kDEA = 0.314 hr-1 TABLE B.16 RUN 16 : 40WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.00 4.000 — _ _ 1 .00 2.800 0.0100 0. 170 0.720 2.08 2.000 0.0720 0.210 1 .230 4.00 1.410 0.2090 0.228 1 .268 5.00 1 . 180 0.2900 0.225 1 .093 6.08 1 .000 0.3850 0.225 0.946 7.17 0.921 0.4860 0.220 0.760 8.08 0.804 0.5400 0.220 0.718 kDEA = 0.320 hr-1 TABLE B.17 RUN 17 : 20WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 2.050 _ _ 1 .00 1 .680 0.0050 0.153 -2.00 1 .280 0.0363 0. 176 0.060 3.00 1 .050 0.0621 0. 186 0.252 4.00 0.858 0.1140 0. 1 93 0.326 5.00 0.770 0.1500 0. 175 0.350 6.00 0.658 0.1780 0. 175 0.291 7.00 0.588 0.2150 0. 176 0.278 8.00 0.538 0.2400 0.181 0.265 kDEA = 0.241 hr-1 TABLE B.18 RUN 18 : 15WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.00 1 .500 — _ 0.50 1 .333 - 0.045 -1 .00 1.212 0.0025 0.065 0.018 2.00 1.161 0.0108 0. 105 0.043 3.00 0.984 0.0280 0.125 0. 186 4.00 0.810 0.0480 0.151 0.242 5.50 0.735 0.0845 0. 152 0.272 6.25 0.650 0.1130 0. 150 0.261 7.56 0.615 0.1280 0.113 0.229 kDEA = 0.131 hr-1 TABLE B.19 RUN 19 : 10WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 0.935 — _ 1 .00 0.820 - 0.073 -2.00 0.735 0.0080 0.125 -3.00 0.643 0.0145 0.140 0.009 4.00 0.575 0.0250 0. 153 0.021 5.00 0.540 0.0495 0. 169 0.031 6.00 0.448 0.0650 0. 163 0.040 7.00 0.387 0.0793 0. 130 0.046 8.00 0.358 0.0838 0.115 0.046 kDEA = 0.104 hr-1 TABLE B.20 RUN 2 0 : 5 WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 0.520 - — _ 0.66 0.470 - 0.020 -1 .75 0.440 - 0.035 -2.75 0.406 0.0025 0.048 0.015 4.08 0.371 0.0049 0.050 0.028 5.08 0.334 0.0105 0.061 0.044 6.08 0.311 , 0.0190 0.074 0.068 7.08 0.300 0.0228 0.063 0.063 8.00 0.260 0.0300 0.055 0.062 kDEA = 0.098 hr-1 TABLE B:21 RUN 21 : 100WT% DEA, 175C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 10.00 — _ _ 1.0 9.0 0.020 0.315 0.810 3.0 7.10 0.092 0.445 2.545 4.0 6.90 0.130 0.435 2.960 5.0 6.30 0. 174 0.415 3.410 6.5 5.50 0.272 0.416 3.050 8.0 4.80 0.334 0.403 2.550 kDEA = 0.092 hr-1 TABLE B.22 RUN 22 : 60WT% DEA, 175C, 4137. kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 6.00 - — _ 1.0 5.35 - 0. 190 0.300 2.0 4.81 0.035 0.360 0.940 3.0 4.25 0.065 0.450 1 .400 4.0 3.74 0.092 0.450 1 .660 5.0 3.38 0.115 0.450 1 .840 6.0 3.21 0.180 0.420 1 .914 7.0 2.65 0.206 0.431 1 .980 8.0 2.25 0.250 0.398 2.046 kDEA = 0.118 hr-1 CSl CN CM O Q O EE .— x •rH E-« Ul a ro o I o O to i— *— K OD ro U w Cu CJ X M \ LO w ro o X *• • W o a X in z n r- o u w Q dp O CM ro CM • « ro CM W PQ Z Q w OH »H 2 rC < to ooooooom in m CM »f o o> r ^-CMro^inininvo oooooooo oooooooo or--«-inintor^cn '-T-CNCNCNCNCNCN OOOOOOOO <*o *t <o in CM ro in r- cn o o o o o o «-I i i o o o o o o cN«-toinm**inroto ocotoincocN — ocn CM i u rC ooooooooo o — cNro**mvor-co CO a CM O CJ a o o to CO OH ro U in < a in CM PQ m < E-CM z D Q o o o o CM in o U in o CM r- CM in U O «—«—«—•— CN CN X 1 EH O O O O O O O CO 1 o *— tootococo«-o — K P CMinr-cn^rovor^-O OOOO"— >— u U 1 u \ X oooooooo to u J o X to ro o co cn co OH o — CN CM in • u o o o o o o u X 1 z PQ o o o o o o o u o CN r- — CO CN CO — < in »jfnmcMr--oo w Q w J a, iH o oooooooo 2 xs, < o r-cNm*invocoo to TABLE B.25 RUN 25 : 10WT% DEA, 175C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 1 .010 — — _ 2.0 0.940 - 0.020 -4.0 0.900 - 0.040 -6.0 0.855 - 0.060 0.048 8.0 0.810 0.0080 0.080 0.068 10.0 0.770 0.0175 0.090 0.060 15.0 0.680 0.0400 0. 100 0.100 20.0 0.610 0.0460 0. 106 0. 142 kDEA = 0.0242 hr-1 TABLE B.26 RUN 26 : 60WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 6.00 _ _ _ 2.0 5.66 - 0.110 0.210 4.0 5.30 - 0.230 0.400 6.0 5.00 0.009 0.382 0.650 8.0 4.71 0.011 0.460 0.850 10.0 4.46 0.021 0.541 1 .000 15.0 3.83 0.043 0.625 1 .380 24.0 2.95 0.110 0.550 1 .520 kDEA = 0.0297 hr-1 TABLE B.27 RUN 27 : 20WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 2.16 — — _ 5.0 1 .92 - 0.100 0.120 10.0 1 .72 - 0.210 0.210 15.0 1 .54 - 0.330 0.300 20.0 1 .39 0.020 0.390 0.380 30.0 1.12 0.033 0.440 0.481 40.0 0.89 0.045 0.470 0.561 60.0 0.65 0.077 0.520 0.632 71.5 0.61 0.089 0.425 0.551 80.0 0.55 0.097 0.440 0.540 97.5 0.53 0. 100 0.410 0.410 kDEA = 0.022 hr-1 TABLE B.28 RUN 28 : 15WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 1 .51 — _ _ 5.0 1 .45 - 0.040 0.024 10.0 1 .35 - 0.082 0.063 15.0 1 .29 0.0040 0.120 0.100 20.0 1.21 0.0060 0.151 0. 132 30.0 1 .09 0.0150 0.215 0.181 40.0 0.98 0.0220 0.221 0.238 50.0 0.92 0.0311 0.225 0.275 kDEA = 0.0104 hr-1 TABLE B.29 RUN 29 : 10WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 1 .000 — — -5.0 0.978 - - -10.0 0.946 - 0.027 -15.0 0.921 - 0.046 -20.0 0.896 - 0.058 -30.0 0.848 0.004 0.086 0.055 40.0 0.803 0.008 0.110 0.085 50.0 0.715 0.012 0. 125 0.092 kDEA = 0.0055 hr-1 TABLE B.30 RUN 30 : 5WT% DEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 0.520 - - -5.0 0.500 - - -10.0 0.488 - 0.0142 -15.0 0.482 - 0.0200 -20.0 0.471 - 0.0294 0.021 30.0 0.440 0.0041 0.0421 0.035 40.0 0.413 0.0053 0.0541 0.039 kDEA = 0.00518 hr-1 TABLE B.31 RUN 31 : 100WT% DEA, 120C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 10.0 - - -19.67 9.54 - 0.580 -43.70 8.77 - 0.992 1 . 160 67.70 7.98 0.0180 1 .063 2.610 91 .70 6.95 0.0425 1 .094 3.050 120.70 5.01 0.0738 1.171 3.351 163.70 3.71 0.0975 1 .170 3.548 187.70 3.21 0.1340 1 . 164 3.544 211.41 2.78 0.1610 1.171 3.320 236.44 2.41 0.1800 1 .175 3.030 kDEA = 0.003 hr-1 TABLE B.32 RUN 32 : 20WT% DEA, 120C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 2. 150 — — — 20.0 2.070 - 0.060 -40.0 2.040 - 0.150 0.050 60.0 1 .930 - 0.210 0.066 80.0 1 .850 - 0.275 0. 104 100.0 1 .710 0.0018 0.330 0.121 140.0 1 .650 0.0033 0.430 0. 154 160.0 1 .538 0.0043 0.450 0. 170 180.0 1 .551 0.0056 0.470 0. 180 200.0 1.510 0.0066 0.460 0.210 kDEA = 0.0022 hr-1 TABLE B.33 RUN 33 : 30WT% DEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.050 — _ _ 1.0 2.290 0.0375 0.150 0.221 2.0 1.810 0.0510 0.210 0.354 3.0 1 .370 0.1060 0.210 0.460 4.0 0.940 0.1560 0.190 0.461 5.0 0.760 0.2320 0.200 0.440 6.0 0.710 0.2950 0.200 0.460 8.0 0.600 0.3710 0. 175 0.450 10.0 0.490 0.4252 0. 150 0.400 13.0 0.410 0.5250 0.150 0.321 24.0 0.203 0.6151 0.100 0. 1 54 27.0 0.200 0.6500 0.121 0.110 31.0 0.1 36 0.7000 0.125 0.091 51 .0 0.075 0.7150 0.115 0.008 kDEA = 0.3 hr-1 TABLE B.34 RUN 34 : 30WT% DEA, 195C, 6895 kPa (1000 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.02 _ _ _ 1 .00 2.38 0.021 0.400 0.205 4.00 1 .28 0.131 0.395 0.850 5.25 0.95 0.179 0.394 0.780 5.93 0.77 0.221 0.375 0.780 6.83 0.62 0.249 0.355 0.750 8.00 0.56 0.272 0.375 0.654 kDEA = 0.23 hr-1 TABLE B.35 RUN 35 : 30WT% DEA, 195C, 5156 kPa (800 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.00 - - -0.67 2.71 0.0275 0.225 0. 150 1 .67 2.05 0.0353 0.375 0.450 2.67 1.71 0.0800 0.388 0.668 3.67 1 .26 0.1110 0.390 0.881 4.67 0.98 0.1710 0.384 0.849 5.67 0.89 0.2060 0.384 0.765 7.17 0.64 0.2660 0.375 0.651 8.00 0.51 0.2752 0.355 0.575 kDEA = 0.23 hr-1 TABLE B.36 RUN 36 : 30WT% DEA, 195C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.040 - - -1 .00 2.380 0.015 0.202 0.210 1 .83 1 .950 0.031 0.247 0.531 3.08 1 .480 0.080 0.250 0.850 4.00 1 .205 0.126 0.228 0.891 5.08 0.907 0. 155 0.258 0.785 7.25 0.689 0.254 0.258 0.714 7.92 0.638 0.267 0.260 0.681 kDEA = 0.23 hr-1 TABLE B.37 RUN 37 : 30WT% DEA, 195C, 3448 kPa (500 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.000 - - -1.0 2.438 0.0055 0.128 0.076 2.0 2. 188 0.0288 0.198 0.250 4.0 1 .630 0.0865 0.190 0.560 5.0 1 .275 0.1220 0.208 0.609 6.0 1 .000 0.1560 0.200 0.660 7.0 0.935 0.2030 0.215 0.786 8.8 0.775 0.2420 0.202 0.761 kDEA = 0.185 hr-1 TABLE B.38 RUN 38 : 30WT% DEA, 195C, 2758 kPa (400 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.030 - - -0.92 2.610 - 0.075 0.074 2.00 2.380 0.0135 0.115 0. 100 3.00 1 .940 0.0475 0.145 0.380 4.08 1 .780 0.0640 0. 138 0.620 5.08 1 .450 0.0840 0.165 0.714 6.00 1 .220 0.1150 0.155 0.721 7.08 1 .050 0.1560 0.145 0.691 8.17 0.985 0.1700 0.148 0.720 kDEA = 0.154 hr-1 TABLE B.39 RUN 39 : 30WT% DEA, 195C, 2069 kPa (300 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.13 _ _ _ 1 .08 2.75 - 0.045 -2.00 2.48 0.0090 0.081 0.051 3.58 2.24 0.0323 0.088 0.247 4.17 1.91 0.0404 0.082 0.384 5.33 1 .88 0.0617 0.099 0.545 6. 17 1 .69 0.0910 0.088 0.650 8.00 1.41 0.1060 0.098 0.785 kDEA = 0.098 hr-1 TABLE B.40 RUN 40 : 30WT% DEA, 195C, 1517 kPa (220 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.10 _ _ 1 .00 2.94 0.0030 0.020 0.021 2.09 2.73 0.0080 0.041 0.043 3.00 2.55 0.0138 0.053 0. 145 4.25 2.45 0.0275 0.054 0.265 6.23 2.25 0.0374 0.053 0.451 8.00 1 .96 0.0428 0.025 0.498 kDEA = 0.061 hr-1 TABLE B.41 RUN 41 : 30WT% DEA, 205C, 41*37 kPa (600 psi) C02 pH adjusted to 12.3 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.000 — — 0.75 2.210 0.0425 0.180 0.248 1 .83 1 .500 0.1080 0. 164 0.650 2.83 1 .060 0.1610 0.171 0.750 3.83 0.815 0.2150 0. 158 0.630 4.83 0.630 0.2640 0. 138 0.710 5.50 0.540 0.3480 0. 125 0.680 7.00 0.478 0.4020 0.078 0.650 8.00 0.428 0.4280 0.071 0.640 kDEA = 0.366 hr-1 TABLE B.42 RUN 42 : 30WT% DEA, 205C, 4137 kPa (600 psi) C02 pH adjusted to 10.0 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 2.93 _ _ 1.0 2.53 - 0. 150 0. 135 2.0 2.35 0.0485 0. 195 0.293 3.0 1 .78 0.1040 0.263 0.504 4.0 1 .50 0.1450 0.258 0.485 5.0 1.41 0.2020 0.242 0.381 6.0 1.19 0.2411 0.245 0.364 8.4 1 .02 0.3980 0. 175 0.228 kDEA = 0.157 hr-1 TABLE B.43. RUN 43 : 30WT% DEA, 205C, 4137 kPa (600 psi) C02 pH adjusted to 9.0 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.02 _ _ _ 1.0 2.78 - 0. 160 0.094 2.0 2.70 0.0125 0.204 0.193 3.0 2.68 0.0425 0.284 0.261 4.0 2.21 0.0925 0.280 0.340 5.0 2.16 0. 1300 0.284 0.300 6.0 2.04 0.1950 0.288 0.280 7.0 1 .92 0.2150 0.304 0.225 kDEA = 0.0675 hr-1 TABLE B.44 RUN 44 : 30 WT%DEA, 205C, 4137 kPa (600 psi) N2 0.14 g/cc K2C03 No degradation took place. TABLE B.4 5 RUN 45 : 30WT% DEA, 205C, 4137 kPa (600 psi) N2 0.1227 g/cc of KHC03 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.10 - - — 2.0 2.50 - 0.062 0.550 4.0 2.02 0.0413 0.075 0.806 6.0 1.61 0.0750 0.075 0.910 8.0 1 .35 0.1350 0.065 0.922 10.0 1 .05 0.1560 0.045 0.910 12.0 0.86 0.2010 0.025 0.840 24.0 0.55 0.3860 - 0.571 28.0 0.47 0.4240 - 0.502 31.0 0.44 0.4322 - 0.446 kDEA = 0.109 hr-1 237 TABLE B.46 RUN 46 : 30WT% DEA, 175C, 4137 kPa (600 psi) N2 0. 1 334 g/cc of KHC03 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3. 12 - — _ 1.0 3.00 - 0.018 0.012 2.0 2.88 - 0.031 0.048 3.4 2.76 0.0035 0.036 0.156 7.0 2.48 0.0075 0.040 0.350 13.0 2.18 0.0148 0.037 0.550 24.0 1 .46 0.0650 0.036 0.810 kDEA = 0. 0245 hr-1 'ABLE B.47 RUN 47 : 30WT% DEA, 150C, 4137 kPa (600 psi) N2 0.218 g/cc of KHC03 SAMPLE CONC. MOLES/CC XI 0-3 hr DEA BHEP HEOD THEED 0.0 3.10 - — _ 6.0 2.98 - 0.0123 0.030 24.0 2.80 - 0.0231 0.246 32.8 2.61 0.0028 0.0250 0.261 47.0 2.38 0.0073 0.0250 0.440 kDEA = 0.0054 hr-1 TABLE B.48 RUN 48 : 66.7WT% DEA in MDEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 6.67 - — — 1 .00 5.80 0.010 0. 130 0.410 2.00 5.20 0.028 0.175 1 .000 4.33 3.10 0.114 0.218 2.060 5.00 2.79 0.143 0.230 2.180 7.37 1 .86 0.268 0.213 2.100 8.75 1 .42 0.296 0.210 1.910 kDEA = 0.1733 hr-1 238 TABLE B.49 RUN 49 : 40WT% DEA in MDEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.92 - _ _ 1.0 3.26 - 0.100 0.302 3.0 2.34 0.0175 0. 175 0.880 4.0 2.03 0.0398 0.201 1.210 5. 1 1 .58 0.0605 0.214 1.418 6. 1 1.41 0.0825 0.216 1 .510 8.0 1 .00 0.1390 0.210 1 .290 kDEA = 0.169 hr-1 TABLE B.50 RUN 50 : 30WT% DEA in MDEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.10 — — _ 1 .00 2.80 - 0.070 0.036 2.00 2.34 - 0. 125 0.151 3.00 2.20 - 0.141 0.212 4. 10 1.71 0.0125 0. 178 0.460 5.00 1 .62 0.0215 0.188 0.571 6.16 1 .28 0.0363 0.202 0.684 7.16 1 .09 0.0563 0.208 0.775 7.84 1 .03 0.0704 0.200 0.860 kDEA = 0.145 hr-1 239 TABLE B.51 RUN 51 : 30 WT% DEA in MDEA, 175C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.000 — _ _ 2.0 2.580 - 0.091 0. 100 4.0 2.214 - 0.201 0.210 6.0 1.911 0.0263 0.271 0.490 8.0 1.614 0.0514 0.304 0.580 kDEA = 0.076 hr-1 TABLE B.52 RUN 52 : 30WT% DEA in MDEA, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 '3.09 _ 5.0 2.82 - 0.040 -27.3 1 .96 - 0. 125 0. 1 46 51.5 1 .32 0.0213 0.238 0.520 kDEA = 0.0204 hr-1 TABLE B.53 RUN 53 : 30WT% DEA in MDEA, 205C, 4137 kPa (600 psi) N2 No degradation took place over a period of 8 hr. TABLE B.54 RUN 54 : 30WT% DEA, 205C, 4137 kPa (600 psi) N2 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.00 2.96 - — _ 7.50 2.83 0.008 - 0.040 24.00 2.77 0.018 - 0.120 31 .50 2.66 0.033 - 0.210 48.00 2.52 0.053 - 0.288 74.50 2.23 0.090 - 0.420 96.00 2.14 0.140 - 0.500 105.55 1 .95 0. 167 - 0.530 130.05 1 .85 0.220 - 0.520 155.00 1 .66 0.247 - 0.500 168.50 1 .60 0.283 - 0.481 179.00 1 .56 0.305 - 0.440 199.04 1 .50 0.345 0.400 kDEA = 0.00365 hr-1 TABLE B.55 RUN 55 : 30WT% DEA, 250C, 4137 kPa (600 psi) N2 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.00 — — — 2.00 2.74 0.030 - 0.020 5.50 2.55 0.075 - 0.113 7.83 2.31 0. 122 - 0. 154 25.33 1 .31 0.492 — 0.221 kDEA = 0.058 hr-1 241 TABLE B.56 RUN 56 : 30WT% DEA, 205C, 8382 kPa (1200 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 3.00 — _ _ 1 .00 2.16 0.0198 0.201 0.530 3.17 1 .08 0.1600 0.280 0.604 4.00 0.79 0.2060 0.261 0.531 5.25 0.53 0.3252 0.275 0.441 6.00 0.41 0.3440 0.280 0.380 7.00 0.31 0.4080 0.228 0.330 8.50 0.23 0.4963 0.223 0.275 kDEA = 0.328 hr-1 TABLE B.57 RUN 56 : 5X10-4 MOLES/CC BHEP, 205C, 4137 kPa (600 psi) C02 No degradation observed over a period of 8 hr . TABLE B.58 RUN 58 : 30WT% DEA + 5X10-4 MOLES/CC BHEP, 205C, 4137 kPa (600 psi) N2 No degradation observed over a period of 8 hr. 242 TABLE B.59 RUN 59 : 30WT% DEA + DEGRADATION PRODUCTS, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.0 3.08 0.355 0.275 0.330 1.0 2.34 0.386 0.283 0.810 2.0 1 .88 0.518 0.315 1 . 160 3.0 1 .55 0.535 0.301 1.131 4.0 1 .38 0.645 0.273 0.861 5.0 1.12 0.754 0.272 0.856 6.0 0.95 0.780 0.253 0.801 7.0 0.86 0.801 0.263 0.850 8.0 0.75 0.836 0.235 0.605 kDEA = 0.26 hr-1 TABLE B.60 RUN 60 : 30WT% DEA + 5X10-4 MOLES/CC BHEP , 205C, (600 psi) C02 SAMPLE CONC. MOLES/CC XI0-3 hr DEA BHEP HEOD THEED 0.0 3.050 0.466 - — 0-.7 2.520 0.481 0.238 0.131 2.5 1 .762 0.576 0.252 0.531 3.2 1 .580 0.607 0.253 0.551 4.0 1 .350 0.654 0.233 0.652 6.0 1 .000 0.721 0.225 0.509 9.0 0.790 0.794 0.21 1 0.400 22.5 0.363 0.953 0. 134 0.202 26.0 0.265 1 .048 0.143 0.051 29.6 0.205 1 .054 0.108 0.010 49.0 0.141 1.171 0.084 — kDEA = 0.219 hr-1 243 TABLE B.61 RUN 61 : 30WT% DEA + 5X10-4 MOLES/CC BHEP, 150C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 3.000 0.466 _ 10.0 2.230 0.501 0.368 0.171 25.0 1 .301 0.523 0.541 0.754 40.0 0.881 0.556 0.560 1 .042 60.0 0.572 0.584 0.534 1 .021 kDEA = 0.0301 hr-1 TABLE B.62 RUN 62 : 3X10-3 MOLES/CC HEOD, 205C, 4137 kPa (600 psi) N2 After 1 hr the analysis showed the presence of HEOD,DEA,THEED, and a trace of BHEP. TABLE B.63 RUN 63 : 10WT% DEA + 0.425X10-3 MOLES/CC HEOD, 205C, 4137 kPa (600 psi) N2 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.0 1 .040 - 0.425 2.0 1 .000 0.0113 0. 178 0.091 5.0 0.938 0.0235 0.083 0. 187 8.0 0.825 0.0275 0.040 0.238 244 TABLE B.64 RUN 64 : 15WT% DEA + DEGRADATION PRODUCTS, 175C, 4137 kPa (600 psi) C02, SAMPLE hr DEA CONC. MOLES/CC X10-3 BHEP HEOD THEED 0.00 1.42 0.0275 0.398 0.970 1 .00 1 .44 0.0463 0.413 1.011 2.23 1.31 0.0701 0.435 1.118 3.00 1.19 0.0825 0.434 1 . 150 5.00 1.10 0.. 1 1 60 0.450 1.181 6.23 1 .04 0.1400 0.442 1.171 7.00 1 .00 0.1580 0.444 1 .063 8.00 0.97 0. 1740 0.448 0.984 TABLE B.65 RUN 65 : 15WT% DEA + DEGRADATION PRODUCTS, 175C, 4137 kPa (600 psi) N2 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 1 .52 0.0275 0.398 0.970 1 .00 1 .50 0.0428 0.285 1 .041 1 .57 1 .46 0.0450 0.253 1.101 3.00 1 .43 0.0664 0.235 1 . 160 4.00 1 .32 0.0780 0.201 1 .202 5.00 1 .29 0.0963 0.193 1 .225 6.12 1 .24 0.1130 0.180 1 .250 7.00 1 .20 0.131 1 0.188 1 .244 8.42 1.18 0.1580 0.191 1 .252 245 TABLE B.66 RUN 66 : 2.6X10-3 MOLES/CC THEED, 205C, 4137 kPa (600 psi) C02 After 1 hr the only product produced was BHEP (-0.6X10.3 moles/cc). TABLE B.67 RUN 67 : 2.6X10-3 MOLES/CC THEED, 205C, 4137 kPa (600 psi) N2 After 1 hr a trace amount of BHEP was produced (-0.12X10-3 moles/cc). TABLE B.68 RUN 68 : 12WT% DEA + 2.6X10-3 MOLES/CC THEED, 205C, 4137 kPa (600 psi) N2 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 1 .210 - - 2.580 1 .00 1 .200 0.210 2.450 2.67 1 .204 0.440 2. 160 4.00 1.212 0.565 2.080 5.75 1 .208 0.670 1 .860 8.33 1.211 0.861 1 .630 246 TABLE B.69 RUN 65 : 12WT% DEA + 2.6X10-3 MOLES/CC THEED, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr DEA BHEP HEOD THEED 0.00 1 . 180 - _ 2.580 1 .00 1.151 0.371 0.285 2.021 2.00 1.111 0.673 0.330 1 .560 3.75 1 .068 1.018 0.233 1 .061 6.33 0.971 1 .260 0. 185 0.860 7.67 0.875 1 .450 0.203 0.731 TABLE B.70 RUN 70 : 30WT% MEA, 205C, 4137 kPa (600 psi) C02 SAMPLE CONC. MOLES/CC X10-3 hr MEA HEI* HEED* 0.0 5.490 - _ 1.0 4.689 80.6 1 .3 2.0 4.115 441 .6 10.5 3.0 3.689 556. 1 12.5 4.0 3.230 698.8 37.7 5.0 2.670 850.4 40.5 6.0 2.246 1028.0 44.8 7.0 2.082 1283.0 83.8 8.0 1 .705 1233.0 110.4 *Conc. as peak area Trace amounts of ozd detected 247 TABLE B.71 RUN 71 : 30WT% TEA, 205C, 4137 kPa (600 psi) C02 No degradation observed over a period of 8 hr. TABLE B.72 RUN 72 : 10WT% MEA + 25WT% DEA, 205C, 4137 kPa(600 psi) C02 SAMPLE CONC . MOLES/CC X10-3 hr MEA DEA BHEP HEOD THEED 0.0 1 .639 2.540 - — _ 0.5 1 .540 2.520 - 0.063 -1.0 1.210 2.050 0.015 0.101 0.108 2.3 0.967 1 .700 0.043 0.124 0.203 3.0 0.803 1 .480 0.1 36 0. 145 0.400 4.0 0.672 1 .200 0.202 0. 125 0.400 5.0 0.541 1 .060 0.281 0. 125 0.401 6.0 0.451 0.841 0.318 0. 120 0.379 7.0 0.410 . 0.800 0.374 0. 120 0.362 23.5 0.230 0.530 0.389 0.095 0. 188 hr HEP* BHEEP* HEI* BHEI * 0.0 - - — _ 0.5 - 104.7 - -1.0 - 355.0 - -2.3 3.3 497.3 49.0 21.0 3.0 37.2 656.2 72.7 141.0 4.0 40.0 617.1 91.0 182.0 5.0 51.5 580.9 127.0 108.0 6.0 63.0 502.3 120.0 137.0 7.0 64.9 487.3 129.0 164.4 23.5 89.0 282.7 193.6 192.0 *Conc. as peak area 248 TABLE B.73 RUN 73 : 1 0WT% TEA + 20WT% DEA, 205C, 4137 kPa(600 psi) C02 SAMPLE CONC. MOLES/CC X10 -3 hr TEA DEA BHEP HEOD THEED TEHEED* 0.0 0.671 2.00 — _ 2.0 0.670 1 .08 -- 0.133 0.230 10.0 4.0 0.662 0.84 0.113 0. 135 0.510 21 . 1 6.0 0.654 0.47 0.250 0. 134 0.433 35.0 8.0 0.649 0.41 0.282 0.131 0.301 51 .4 *Conc. as peak area 249 APPENDIX C The solubility of C02 in DEA solutions at high temperature and pressure Some knowledge of the solubility of C02 under operating conditions was required in order to carry out the ionic experiments in section 9.2. Unfortunately the open literature did not provide the required information. Therefore, a series of simple solubility experiments were performed to provide the necessary data. C.1 Experimental method Essentially the method consisted of: 1. Filling a 2 L high pressure bomb to 5156 kPa (800psi) with C02 and then weighing the bomb and C02. The scales were accurate for changes of down to 0.2 g. 2. The 600 ml autoclave (see sect. 6.1) was charged with 450 ml of a specified aqueous DEA solution and sealed. The autoclave was then heated to the required temperature with the stirrer in operation at about 150 r.p.m. 3. The bomb was then connected to the autoclave and C02 was fed into the autoclave to the required pressure. After equilibrium had been reached the bomb was disconnected. 4. The bomb was then reweighed and the weight of C02 fed to the auto clave noted. 5. The above procedure was repeated many times to cover the following range of conditions: a) DEA concentration:- 30, 20, and 10 wt % b) Temperature:- 205-100°C c) Overall pressure: 413.7-4137 kPa (60-600 psi) It should be noted that overall pressure in the autoclave was made up 250 from the partial pressure of C02 and the vapour pressure of the water vapour from the aqueous DEA solution and to a very small extent from the DEA itself. For example at 205°C the vapour pressure of pure water is 1700 kPa (246.6 psi) and the vapour pressure of DEA is 8.6 kPa (1.25 psi). Therefore to determine the partial pressure of C02 in the auto clave a knowledge of the vapour pressure of the aqueous DEA solution was required. To determine the vapour pressure the following simple experiments were performed. A specified concentration of DEA was charged to the autoclave. The autoclave was heated to a specified temperature and the pressure noted. The temperature was then raised and the pressure noted again. This was repeated up to 205°C. Table C.l gives the results which are plotted in Figure C.l. Table C.l Vapour pressure of DEA solutions as a function of temperature Vapour pressure psi Temp. °C 0 psi kPa DEA 10 psi cone kPa wt % 20 psi kPa 30 psi kPa 100 14.69 101 .3 14.69 101. 3 14.69 101 .3 14.69 101 .3 120 29.40 202 .7 28.69 197. 8 27.0 186 .2 24.43 168 .4 140 54.96 379 .0 52.57 362. 5 49.1 338 .5 43.57 300 .4 160 92.50 637 .8 88.14 607. 7 82.29 567 .4 71.93 496 .0 180 148.96 1027 .1 142.86 985. 133.29 919 .0 121.43 837 .3 200 227.50 1568 .6 221.29 1525. 8 214.29 1477 .5 196.14 1352 .4 210 285.9 1971 .3 273.43 1885. 3 259.71 1790 .7 243.29 1677 .5 251 TEMPERATURE (°C) Figure C.l DEA solution vapour pressure as a function of temperature and DEA concentration 252 C.2 Calculation of CO, solubility From the knowledge of the total weight of C02 used, total pressure in reactor, vapour pressure of aqueous DEA solution, temperature and concentration of DEA, the solubility of C02 was calculated as g of C02/g of DEA using the following method. The main problem was to determine the mass of C02 in the vapour phase. It was first assumed the mixture of C02, water vapour, and DEA vapour was ideal and the partial pressure of each component was propor tional to its molar concentration. Therefore to determine the partial pressure of C02 in the vapour phase the vapour pressure of the aqueous DEA solution was simply subtracted from the total pressure, i.e., PP = P -V.P. fC.ll r-r,C02 total DEA solution 1 J Knowing the partial pressure of C02, the volume of the vapour phase (i.e., 600-450 ml = 150 ml) and the temperature; the number of moles of C02 could be calculated using an equation of state. Since for the range of conditions studied the compressibility factor for C02 lay in the range 0.99-0.92 it could be assumed C02 existed as an ideal gas. j^ie yan ^er Waals equation of state was used to calculate the number of moles. (P + (V-nb) = nRT [C.2] Although there are more accurate equations of state available, the Van der Waals relation is still useful for providing an approximate yet simple, analytical representation of the behaviour of a gas. Once the number of moles of C02 in the vapour phase had been deter mined the mass of C02 in the DEA solution could simply be obtained by subtracting the vapour phase mass of C02 from the total mass of C02 fed to the autoclave, i.e., 253 Mass of C02 dissolved in DEA solution = total mass of C02 fed to autoclave - mass of C02 in vapour phase [C.3] From the mass of C02 dissolved in the aqueous DEA solution the solubility •of C02 as g C02/g DEA could be easily calculated. C.3 Example The following shows how the solubility of C02 in a 30 wt % DEA solution at 205°C under a total pressure of 4137 kPa (600 psi) was cal culated . Total weight of C02 fed to autoclave = 31.5 g Volume of Reactor = 625 cc Volume of Solution = 410 cc Volume of C02 = 215 cc Density of 30 wt % DEA = 1.088 Wt of DEA in solution = 133.9 g a) At 205°C the vapour pressure of 30 wt % DEA = 1510 kPa (219 psi) the partial pressure of C02 in the vapour phase = 4137 - 1510 = 2627 kPa (381 psi) b) 2627 kPa (381 psi) of C02 at 205°C in a volume of 215 cc corresponds to a mass of n moles using Eqn. C.2. P = 2627 kPa (381 psi) = 25.94 atm V  0.215 L T = 205°C = 478°K R = 0.082055 L atm/°K mole For C02:-a = 3.59 L2 atm/mole2 b = 0.0427 L/mole 254 Eqn. C.2 becomes (25.94 + "p I'^y) (0.215 - n(0.0427)) = n x 0.082055 * 478 n = 0.151 moles = 6.47 g c) The mass of C02 dissolved in the DEA solution is: 31.5 - 6.47 = 25.03 g d) Therefore the solubility of C02 is:-= 0.187 g C02/g DEA It is realized that the calculation of the mass of C02 in the vapour phase may be somewhat inaccurate. However, when it is noted that the mass of C02 in the vapour phase is usually less than 30% of the C02 dis solved in the liquid phase then errors of about ± 24% in the vapour phase C02 mass will only cause an error of about ± 10% in the calculation of solubility. An accuracy of ± 10% is considered suitable for these studies. C.4 Results The following tables give the solubility of C02 in DEA solutions at varying temperature and total pressure as a function of C02 partial pressure in the vapour phase. Figures 9.2 to 9.4 in section 9.2 summar ize these tabulated results. TABLE C.2 SOLUBILITY OF C02 IN 30WT% DEA TEMP. TOTAL PRESSURE PARTIAL PRESSURE SOLUBILITY OF C02 OF C02 C psi kPa psi kPa gC02/gDEA 200 550.0 3793.3 400.7 2762.9 0. 183 405.0 2792.5 208.7 1439.0 0.114 n 275.0 1896.1 78.7 542.6 0.089 190 470.0 3240.7 375.3 2587.7 0.190 n 350.0 2413.3 194.3 1340.0 0.120 ?? 236.0 1627.2 80.3 553.7 0.081 180 400.0 2758.0 350.6 2417.4 0.192 it 295.0 2034.0 173.6 1197.0 0. 123 it 196.0 1351.4 74.6 514.4 0.090 170 520.0 3585.4 427. 1 2945.0 0.236 n 338.0 2330.5 310.1 2138.2 0.202 it 245.0 1689.3 1 52. 1 1048.7 0. 128 it 160.0 1 103.2 67. 1 462.7 0.090 160 452.0 3136.5 379.7 2618.0 0.240 it 290.0 2000.0 250.7 1728.6 0.200 it 195.0 1334.5 122.7 846.0 0. 132 n 132.0 910.1 59.7 411.6 0.094 150 600.0 4137.0 543.4 3746.7 0.312 n 360.0 2482.2 301 .4 2078.2 0.240 n 250.0 1723.8 193.4 1333.5 0. 195 n 1 52.0 1048.0 95.4 657.8 0. 1 37 n 1 05.0 723.9 48.4 333.7 0. 1 04 1 40 508.0 3502.7 464.6 3203.4 0.328 « 291 .0 2006.4 247.6 1707.2 0.248 it 220.0 1516.9 176.6 1217.7 0.209 n 118.0 813.6 74.6 514.4 0.141 it 83.0 572.3 39.6 273.0 0. 106 130 445.0 3068.3 412.7 2845.6 0.330 n 240.0 1654.8 207.7 1432.1 0.249 n 177.0 1220.4 145.0 999.8 0.210 n 90.0 620.6 57.0 393.0 0. 144 n 65.0 448.2 32.7 225.5 0.111 120 555.0 3826.7 530.7 3659.2 0.376 n 410.0 2826.9 385.7 2659.4 0.330 it 200.0 1379.0 175.7 1210.1 0.248 n 140.0 965.3 115.7 797.8 0.212 •t 68.0 468.9 43.7 301 .3 0. 147 n 45.5 313.7 21.2 146.7 0.111 110 510.0 3516.5 491 .4 3388.2 0.381 •t 390.0 2689.1 371 .4 2560.8 0.344 it 170.0 1 172.2 151 .4 1043.9 0.256 it 119.0 820.5 100.4 692.3 0.215 it 53.0 365.4 34.4 237.2 0.149 n 32.6 224.8 14.0 96.5 0.108 100 475.0 3275.1 461 .6 3182.7 0.392 It 355.0 2447.7 341 .6 2355.3 0.355 It 150.0 1034.3 136.6 941 .9 0.264 It 89.0 613.7 75.6 521 .3 0.217 It 40.0 275.8 26.6 183.4 0.149 It 24.0 165.5 10.6 73.1 0.108 TABLE C.3 SOLUBILITY OF C02 IN 20WT% DEA 256 TEMP. TOTAL PRESSURE PARTIAL PRESSURE SOLUBILITY OF C02 OF CO 2 C psi kPA psi kPa gC02/gDEA 200 560.0 3861.2 353.3 2436.0 0. 198 n 405.0 2792.5 198.3 1367.3 0. 152 n 307.0 2116.8 100.3 691 .6 0.112 190 503.0 3468.2 338.2 2331.9 0.200 fi 342.0 2358.1 183.2 1263.2 0. 154 n 246.0 1696.2 87.2 601 .2 0.115 180 436.0 3006.2 305.0 2103.0 0.205 n 286.0 1972.0 1 55. 1 1069.4 0.159 fi 214.0 1475.5 83. 1 573.0 0. 123 170 552.0 3806.0 | 450.5 3106.2 0.255 n 378.0 2606.3 276.5 1906.5 0.211 n 238.0 1641.0 136.5 941 .2 0. 162 fi 180.0 1241.1 78.5 541 .3 0. 129 160 489.0 3371.7 410.0 2827.0 0.270 fi 314.0 2165.0 235.0 1620.3 0.226 fi 190.0 1310.1 111.0 765.3 0. 168 140.0 965.3 61 .0 420.6 0. 142 150 600.0 4137.0 538.5 3713.0 0.315 fi 426.0 2937.3 364.5 2513.2 0.280 256.0 1765.1 194.5 1341 .1 0.236 150.0 1034.3 88.5 610.2 0.183 ti 108.0 744.7 46.5 320.6 0. 144 140 525.0 3619.9 477.7 3293.7 0.330 364.0 2509.8 316.7 2182.7 0.290 217.0 1496.2 169.7 1156.3 0.240 n 115.0 792.9 67.7 466.8 0. 177 fi 82.0 565.4 34.7 239.3 0. 146 130 469.0 3233.8 433.7 2990.4 0.339 n 301 .0 2075.4 265.7 2521.5 0.295 fi 186.0 1282.5 150.7 1039.1 0.248 fi 86.0 593.0 50.7 349.6 0. 183 n 61.0 420.6 25.7 177.7 0. 148 120 516.0 3557.8 490.0 3378.6 0.381 n 410.0 2827.0 384.0 2647.7 0.351 266.0 1834. 1 240.0 1654.2 0.306 n 156.0 1075.6 130.0 896.4 0.257 68.5 472.3 42.5 293.0 0.185 47.0 324.1 21.0 144.8 0. 150 110 470.0 3240.7 450.5 3106.2 0.391 ft 367.0 2530.5 347.5 2396.0 0.357 220.0 1516.9 200.5 1382.5 0.315 137.0 944.6 117.5 810.2 0.270 54.5 375.8 35.0 241 .3 0.190 n 35.0 241 .3 15.5 106.9 0.153 100 414.0 2854.5 400.2 2759.4 0.404 fl 327.0 2254.7 313.2 2159.5 0.365 fl 199.0 1372.1 185.2 1276.9 0.318 f» ,113.8 784.6 100.0 689.5 0.275 fl 42.0 289.6 28.2 194.4 0.197 n 27.0 186.2 13.2 91 .0 0.156 TABLE C.4 SOLUBILITY OF C02 IN 10WT% DEA 257 TEMP. . TOTAL PRESSURE PARTIAL PRESSURE SOLUBILITY psi OF C02 OF CO 2 C kPa psi kPa gC02/gDEA 200 555.0 3826.7 337.9 2329.8 0.279 n 418.0 2882.1 200.9 1385.2 0.246 n 275.0 1896.0 70.9 488.9 0. 185 190 495.0 3413.0 321 .3 2214.7 0.290 it 363.0 2502.9 189.2 1303.2 0.249 n 238.0 1641.0 64.2 442.7 0.189 180 431 .0 2971.7 290.5 2003.0 0.295 ti 310.0 2137.5 169.5 1168.7 0.257 n 198.0 1365.2 57.5 369.5 0.191 170 555.0 3826.7 444.9 3067.6 0.338 fi 380.0 2620.1 269.9 1861.0 0.300 262.0 1806.5 151.9 1047.4 0.262 163.0 1123.9 52.9 364.7 0. 193 160 482.0 3323.4 396.2 2731.8 0.352 n 332.0 2289.1 246.2 1697.5 0.316 ti 218.0 1503.1 132.2 91 1 .5 0.270 n 130.0 896.4 44.2 304.8 0.196 1 50 600.0 4137.0 533.5 3678.5 0.423 ti 416.0 2868.3 349.5 2409.8 0.376 n 281 .0 1937.5 214.5 1479.0 0.334 « 181.0 1248.0 114.5 789.5 0.279 fi 103.0 710.2 36.5 251 .7 0.200 140 533.0 3675.0 481 .9 3322.7 0.464 fi 360.0 2482.2 295.9 2040.2 0.396 w 232.0 1599.6 180.9 1247.3 0.349 n 147.0 1013.6 95.9 661 .2 0.287 II 80.0 551 .6 28.9 199.3 0.200 130 478.0 3295.8 439.6 3031.0 0.482 n 312.0 2151.2 273.6 1886.5 0.408 n 193.0 1330.7 154.6 1066.0 0.355 n 123.4 850.8 85.0 586. 1 0.292 n 62.0 427.5 23.6 162.7 0.205 120 566.0 3902.6 538.3 3711.6 0.558 n 435.0 2999.3 407.3 2808.3 0.510 n 266.0 1834.1 238.3 1645.1 0.430 fi 163.0 1123.9 135.3 932.9 0.370 n 98.0 675.7 70.3 484.7 0.300 47.0 324.1 19.3 133.1 0.200 110 510.0 3516.5 489.6 3375.8 0.580 n 392.0 2702.9 371 .6 2562.2 0.530 It 226.0 1558.3 205.6 1417.6 0.435 fi 136.0 937.7 1 15.6 797.1 0.376 fi 81.0 558.5 60.6 417.8 0.303 it 38.0 262.0 17.6 121.4 0.207 100 463.0 3192.4 448.7 3093.8 0.598 fl 353.0 2434.0 338.7 2334.4 0.540 fl 192.0 1324.0 177.7 1255.2 0.450 fl 120.0 827.4 105.7 728.8 0.382 fl 65.0 448.2 50.7 349.6 0.305 fl 29.0 200.0 15.3 105.5 0.208 258 APPENDIX D Derivation of the kinetic model Using the simplified degradation route developed in section 12.1, i.e., . HEOD THEED BHEP the following equations were derived for the rate of change of the various compounds. dl°"] • -MDEA] - k2[DEA] [D.l] dIDEA] , .WDEA] [B_2] where kD£A = kx + k2 [D.3= kllDEAl [D.4] dt d[T"£ED] = k2[DEA] - k3[THEED] [D.5] dt d[^HEP] = k3[THEED] [D.6] dt Integrating Eqn. D.l yields [DEA]^ = [DEA]o e"(kl+k2)t [D.7] Equation D.4 was then integrated: di§E°5l = ki[DEA] = k1[DEA]0 e~(kl+k2)t dt t u [HEOD]t = k1[DEA]0[- e-(ki+k2)t]t [HEOD]t = [DEA] o (1 - e"(kl+k2)t) [D.8] Equation D.5 was integrated as follows: iLZHEEDl + ic_3 [THEED] - k2[DEA]0 e"0^^ dt t This is a first-order linear differential equation which can be multiplied k, t by the integrating factor e J . [THEED]^ ek3t = k2[DEA]0 J* e<k3-(ki+k2))t dt = k™ flcTTO) e(k-^))t]t [THEED]t = [DEA]0 ^.(j^) (e_(kl+k2)t - e"k3t) [D.9] Equation D.6 was then solved. d[BHEP] _ rTUFFHi - i, rnrM k2 , -<ka+k2)t -k3t. ^- - k3 [THEED] t - k3 [DEA] „ ^Tj^y (e -e ) [BHEP] = [DEA] o . , [~ jr-^ e"(ki+k2)t 1 -kst t t " k3-(ki+k2) (ki+k2) k3 o ii /, .i \ ~k3t , -(ki+k2)t t - rr.FAl k2k3 , (ki+k2) e - k3e 1J [DtAio k3-(k1+k2) 1 k3(k1+kz)J0 This simplifies to: [BHEP] = [DEA] o r^- U " „ <l\^ , e~^+^)t ki+k2 e"k3t t ki+k2 k3-(ki+k2) k3-(k!+k2) [D.10] In many cases the plots of [THEED] versus time pass through a maxi mum. The location of this maxima can be found by differentiating Eqn. D.9 and setting d[THEED]/dt = 0. The time at which the maximum concen tration of THEED occurs is thus: d[THEED] [DEA]Q k2 , . -(ki+k2)t max ^ . -k3t max. . dt = k3-(k1+k2) ("(kl+k2) e +k3* )=0 -(ki+k2)t max , e k3 e~k3t max (ki+k2) The maximum concentration of THEED can be found by combining Eqns. D.9 and D.ll: 260 _ (k2+ki) m , k3 , _ _kj , k3 ^ [THEED] _ [DEA]p k2 r k3-(k24-ki) k2+k/_ k3-(k2+k!) ' "Sti+kA max k3-(ki+k2) J This simplifies to: [THEED] max k2 ,k2+ki k3~(kl+k2) r [DEA]o k2+ki K k3 ; [D>12] 261 APPENDIX E Comparison between the experimental results and the prediction of the  kinetic model The following tables compare the experimentally measured values of DEA, BHEP, HEOD, and THEED concentrations with the values predicted by the kinetic model developed in chapter 12. For all the tables the concentra tion in moles/cc and all k values are given in hr ^. For each case the total operating pressure is 4137 kPa (600 psi). TABLE E.1 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 60WT% DEA DEGRADED AT 175C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 6.000 6.000 — — 1.0 5.350 5.330 - 0.0080 2.0 4.810 4.741 0.0350 0.0321 3.0 4.250 4.211 0.0651 0.0678 4.0 3.741 3.740 0.0920 0.1161 5.0 3.380 3.330 0.1150 0.1718 6.0 3.210 2.960 0.1800 0.2360 7.0 2.650 2.630 0.2060 0.3060 8.0 2.251 2.334 0.2511 0.3820 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - — — -1 .0 0. 190 0.205 0.300 0.454 2.0 0.360 0.380 0.940 0.842 3.0 0.451 0.550 1.400 1.171 4.0 0.450 0.670 1 .661 1 .453 5.0 0.440 0.822 1 .840 1 .682 6.0 0.420 0.935 1.914 1 .870 7.0 0.431 0.950 1 .980 2.030 8.0 0.398 1.010 2. 150 2. 150 kDEA = 0.118 k1 = 0.0363 k2 = 0.0817 k3 = 0.0360 TABLE E.2 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 60WT% DEA DEGRADED AT 150C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 6.000 6.000 _ _ 2.0 5.660 5.650 - -4.0 5.300 5.330 - -6.0 5.000 5.022 0.0090 0.0076 8.0 4.710 4.730 0.0110 0.0101 10.0 4.460 4.460 0.0210 0.0200 15.0 3.830 3.240 0.0430 0.0430 24.0 2.950 2.940 0.1100 0.1010 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - - — _ 2.0 0.110 0. 1 27 0.210 0.218 4.0 0.230 0.247 0.400 0.422 6.0 0.383 0.359 0.650 0.612 8.0 0.460 0.466 0.850 0.790 10.0 0.541 0.566 1 .000 0.956 15.0 0.625 0.792 1 .380 1 .320 24.0 0.550 1.120 1 .520 1 .840 kDEA = 0.0297 ki = 0.0109 k2 = 0.0188 k3 = 0.0040 TABLE E.3 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 30WT% DEA DEGRADED AT 175C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3.040 3.040 _ 1.0 2.670 2.660 - -2.0 2.360 2.360 - 0.0139 3.0 2.000 2.051 0.0150 0.0350 4.0 1 .840 1 .849 0.0280 0.0500 5.0 1 .580 1.610 0.0580 0.0611 6.0 1 .440 1 .450 0.0620 0.1030 8.0 1.110 1 . 1 30 0.0980 0.1700 10.0 0.860 0.895 0.2060 0.2370 12.0 0.651 0.702 0.2200 0.3120 14.0 0.520 0.550 0.2840 0.3810 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ _ 1.0 0. 1 50 0. 1 34 0. 130 0.200 2.0 0.250 0.253 0.280 0.380 3.0 0.360 0.358 0.450 0.525 4.0 0.410 0.451 0.660 0.650 6.0 0.425 0.534 0.760 0.750 8.0 0.485 0.729 0.910 0.970 10.0 0.541 0.825 1 .050 1 .044 12.0 0.550 0.850 0.990 1 .066 14.0 0.540 0.870 0.960 1 .025 kDEA = 0.1210 k1 = 0.0474 k2 = 0.0736 k3 = 0.0350 TABLE E.4 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS FOR 30WT% DEA DEGRADED AT 162C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3.000 3.000 1.0 2.803 2.800 - -2.0 2.620 2.610 - -3.0 2.450 2.410 0.0101 0.0073 4.0 2.280 2.240 0.0150 0.0126 5.0 2. 137 2. 1 20 0.0201 0.0194 6.0 2.000 1 .975 0.0254 0.0270 7.0 1 .870 1 .880 0.0350 0.0354 8.0 1.744 1 .720 0.0420 0.0460 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - - — _ 1.0 0.105 0.075 0.090 0. 120 2.0 0.200 0. 146 0.200 0.230 3.0 0.275 0.212 0.290 0.330 4.0 0.320 0.273 0.388 0.426 5.0 0.364 0.330 0.460 0.513 6.0 0.412 0.385 0.561 0.591 7.0 0.420 0.435 0.653 0.663 8.0 0.450 0.482 0.731 0.730 kDEA = 0.0678 k1 = 0.0260 k2 = 0.0418 k3 = 0.0140 TABLE E.5 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 30WT% DEA DEGRADED AT 150C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3.000 3.000 _ 5.0 2.600 2.600 0.0098 0.0024 10.0 2.200 2.200 0.0210 0.0100 15.0 1 .860 1 .881 0.0380 0.0190 20.0 1 .560 1 .610 0.0425 0.0322 25.0 1.310 1 .380 0.0550 0.0480 30.0 1.110 1.180 0.0640 0.0652 40.0 0.840 0.868 0.0860 0.0980 50.0 0.710 0.637 0.1060 0.1090 60.0 0.580 0.480 0.1300 0.1490 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ _ _ 5.0 0.250 0. 197 0.075 0.231 10.0 0.350 0.366 0. 150 0.424 15.0 0.470 0.501 0.400 0.560 20.0 0.510 0.640 0.652 0.719 25.0 0.550 0.740 0.751 0.830 30.0 0.570 0.830 0.900 0.920 40.0 0.570 0.980 1 .075 1 .059 50.0 0.542 1 .050 1 .225 1 . 132 60.0 0.540 1 .070 1 .200 1 .189 kDEA = 0.0316 k1 = 0.0142 k2 = 0.0168 k3 = 0.0040 TABLE E.6 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 30WT% DEA DEGRADED AT 140C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3.000 3.000 _ 10.0 2.850 2.850 - -20.0 2.540 2.500 0.0120 0.0042 30.0 2.260 2.200 0.0210 0.0110 40.0 2.000 2.000 0.0330 0.0180 50.0 1 .754 1 .745 0.0415 0.0370 60.0 1 .605 1 .560 0.0510 0.0480 80.0 1 .270 1 .240 0.0700 0.0650 100.0 1.010 0.980 0.1050 0.0890 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ _ _ 10.0 0.250 0.171 0. 120 0. 1 76 20.0 0.450 0.322 0.320 0.330 30.0 0.575 0.460 0.504 0.466 40.0 0.660 0.576 0.581 0.584 50.0 0.710 0.700 0.700 0.700 60.0 0.710 0.780 0.751 0.775 80.0 0.660 0.940 0.863 0.921 100.0 0.700 1 .060 0.950 1 .031 kDEA = 0.01150 k1 = 0.00563 k2 = 0.00587 k3 = 0.00140 TABLE E.7 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 30WT% DEA DEGRADED AT 120C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3. 150 3. 150 — _ 20.0 3.000 2.990 - -40.0 2.900 2.840 - -60.0 2.800 2.702 - -80.0 2.650 2.560 - -100.0 2.500 2.430 0.0060 0.0033 120.0 2.340 2.306 0.0084 0.0041 140.0 2.244 2.200 0.0100 0.0063 160.0 2.131 2.080 0.0180 0.0088 180.0 2.000 1 .970 0.0250 0.0121 200.0 1 .905 1 .870 0.0310 0.0151 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — — — — 20.0 0. 150 0. 120 - -40.0 0.300 0.218 0.056 0.093 60.0 0.380 0.318 0.089 0. 125 80.0 0.525 0.414 0. 100 0. 175 100.0 0.600 0.505 0. 140 0.213 1 20.0 0.705 0.590 0.225 0.250 140.0 0.751 0.673 0.240 0.280 160.0 0.760 0.750 0.313 0.313 180.0 0.770 0.824 0.345 0.343 200.0 0.780 0.894 0.375 0.371 kDEA = 0.00260 k1 = 0.00182 k2 = 0.00078 k3 = 0.00030 TABLE E.8 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 30WT% DEA DEGRADED AT 90C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 3.580 3.580 _ _ 100.0 3.520 3.533 - -160.0 3.420 3.487 - — 300.0 3.455 3.441 - — 441 .0 3.403 3.378 - -511.0 3.361 3.346 - — 631 .0 3.310 3.300 - — 700.0 3.280 3.264 - -SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ 1 00.0 0.050 0.046 - — 160.0 0.090 0.092 - — 300.0 0. 155 0. 137 - 0.0181 441 .0 0.191 0. 199 0.018 0.0302 511.0 0.216 0.230 0.024 0.0348 631 .0 0.250 0.282 0.048 0.0427 700.0 0.290 0.310 0.060 0.0507 kDEA = 0.000142 k1 = 0.000140 k2 = 0.000002 k3 = -TABLE E.9 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 20WT% DEA DEGRADED AT 175C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 2.020 2.020 _ 1 .0 1.810 1.810 - -2.0 1 .660 1 .640 - -3.0 1 .550 1 .480 0.024 0.025 4.0 1 .350 1 .340 0.036 0.038 5.0 1 .240 1.210 0.054 0.046 6.0 1 . 150 1 .090 0.074 0.050 7.0 1 .030 0.986 0.096 0.082 8.0 0.910 0.890 0. 105 0.112 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - _ _ „ 1 .0 0. 100 0.061 0. 150 0.129 2.0 0. 170 0.115 0.250 0.241 3.0 0.210 0.165 0.320 0.336 4.0 0.250 0.210 0.440 0.419 5.0 0.250 0.250 0.500 0.489 6.0 0.260 0.286 0.540 0.548 7.0 0.270 0.319 0.590 0.598 8.0 0.290 0.339 0.615 0.638 kDEA = 0.1010 k1 = 0.0318 k2 = 0.0692 k3 = 0.0390 TABLE E.10 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS FOR 20WT% DEA DEGRADED AT 150C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 2. 160 2. 160 _ _ 5.0 1 .920 1 .940 - -10.0 1 .720 1 .730 - -15.0 1 .540 1 .553 - -20.0 1 .390 1 .390 0.020 0.016 30.0 1 . 120 1 . 120 0.033 0.033 40.0 0.890 0.896 0.045 0.055 60.0 0.650 0.580 0.077 0.096 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ _ 5.0 0. 100 0.112 0. 120 0.110 10.0 0.210 0.213 0.210 0.219 15.0 0.330 0.304 0.300 0.294 20.0 0.390 0.384 0.380 0.368 30.0 0.440 0.522 0.481 0.488 40.0 0.470 0.570 0.561 0.577 60.0 0.520 0.791 0.632 0.686 kDEA = 0 .022 k1 0 .011 k2 0 .011 k3 0 .004 TABLE E.11 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 20WT% DEA DEGRADED AT 120C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 2.150 2.1 50 _ 20.0 2.070 2.057 - -40.0 2.040 1 .970 - -60.0 1.930 - 1 .880 - -80.0 1 .850 1 .800 - -100.0 1.710 1 .725 0.0018 0.0020 1 40.0 1 .650 1 .580 0.0033 0.0033 160.0 1 .538 1.513 0.0043 0.0045 180.0 1 .551 1 .450 0.0056 0.0053 200.0 1.510 1 .390 0.0066 0.0063 SAMPLE DEA THEED hr EXP CALC EXP CALC 0.0 _ _ 20.0 0.060 0.067 - 0.025 40.0 0.150 0. 132 0.050 0.049 60.0 0.210 0. 195 0.066 0.072 80.0 0.275 0.252 0. 104 0.088 100.0 0.330 0.309 0.121 0.114 140.0 0.430 0.415 0. 154 0. 152 160.0 0.450 0.460 0. 170 0. 170 180.0 0.470 0.510 ,0.180 0.181 200.0 0.460 0.557 0.210 0.202 kDEA = 0.0022 k1 = 0.0016 k2 = 0.0006 k3 <= 0.0003 TABLE E.12 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 15WT% DEA DEGRADED AT 175C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 1 .500 1 .500 _ 1.0 1 .420 1 .436 - — 2.0 1 .370 1 .369 - -3.0 1.310 1 .294 0.0060 0.0040 4.0 1 .240 1 .241 0.0130 0.0120 5.0 1 . 180 1 . 176 0.0200 0.0208 6.0 1 . 120 1 . 125 0.0280 0.0320 8.0 1 .080 1 .020 0.0490 0.0487 10.0 1.010 0.930 0.0580 0.0618 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - _ 1 .0 0.026 0.025 - -2.0 0.050 0.050 0.050 0.067 3.0 0.076 0.073 0. 100 0. 100 4.0 0.093 0.940 0. 120 0. 122 5.0 0.118 0.118 0. 170 0. 146 6.0 0.131 0. 135 0. 172 0. 1 67 8.0 0. 160 0. 172 0.225 0.204 10.0 0.171 0.205 0.250 0.234 kDEA = 0.0490 k1 = 0.0173 k2 = 0.0317 k3 = 0.0400 TABLE E.13 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 15WT% DEA DEGRADED AT 150C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 1.510 1.510 _ 5.0 1 .450 1 .420 - -10.0 1.350 1 .346 - 0.0016 15.0 1 .290 1 .274 0.0040 0.0038 20.0 1 .210 1.212 0.0060 0.0060 30.0 1 .090 1 .085 0.0150 0.0130 40.0 0.980 0.974 0.0220 0.0218 50.0 0.920 0.874 0.0340 0.0324 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 - — _ _ 5.0 0.040 0.042 0.024 0.055 10.0 0.082 0.081 0.063 0.071 15.0 0. 120 0.118 0. 1 00 0. 100 20.0 0.151 0. 154 0. 132 0. 132 30.0 0.215 0.219 0.181 0. 183 40.0 0.221 0.278 0.238 0.227 50.0 0.225 0.330 0.275 0.263 kDEA = 0.0108 k1 = 0.0057 k2 =0.0051 k3 = 0.0043 TABLE E.14 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS. FOR 10WT% DEA DEGRADED AT 175C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 1 .010 1.010 2.0 0.940 0.953 - — 4.0 0.900 0.900 - — 6.0 0.855 0.865 - — 8.0 0.810 0.824 0.0080 0.0141 10.0 0.770 0.780 0.0175 0.0196 15.0 0.680 0.696 0.0400 0.0400 20.0 0.610 0.610 0.0460 0.0648 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 — _ 2.0 0.020 0.020 - — 4.0 0.040 0.040 - — 6.0 0.060 0.059 0.048 0.070 8.0 0.080 0.077 0.068 0.090 10.0 0.090 0.094 0.060 0.101 15.0 0.100 0. 133 0. 100 0.131 20.0 0.106 0. 168 0. 142 0.151 kDEA = 0.0242 k1 = 0.0106 k2 = 0.0136 k3 = 0.0340 TABLE E.15 COMPARISON BETWEEN EXPERIMENTAL AND PREDICTED CONCS FOR 10WT% DEA DEGRADED AT 150C SAMPLE DEA BHEP hr EXP CALC EXP CALC 0.0 1 .000 1 .000 5.0 0.978 0.974 - — 10.0 0.946 0.940 - _ 15.0 0.921 0.920 - _ 20.0 0.896 0.894 - 0.0018 30.0 0.848 0.850 0.0040 0.0040 40.0 0.803 0.804 0.0080 0.0071 50.0 0.715 0.760 0.0120 0.0106 SAMPLE HEOD THEED hr EXP CALC EXP CALC 0.0 5.0 - - — 10.0 0.027 0.029 — _ 15.0 0.046 0.043 - — 20.0 0.058 0.057 - 0.0456 30.0 0.086 0.083 0.0550 0.0600 40.0 0.110 0.108 0.0851 0.0827 50.0 0. 125 0.131 0.0920 0.0987 kDEA = 0.0055 k1 = 0.0030 k2 = 0.0025 k3 = 0.0040 277 APPENDIX F Mass spectra of DEA and related degradation compounds The following is basically a library of mass spectra for DEA and its degradation compounds. Out of the 16 compounds analysed only spectra for MEA, DEA, and TEA could be found in existing mass spectra libraries. F.1 Mass spectra of DEA and its degradation compounds Figures F.l to F.4 show the mass spectra of DEA, BHEP, HEOD, and THEED. F.2 Mass spectra of minor degradation compounds Figures F.5 to F.ll show the mass spectra of BHEED, HEED, HEI, HEM, HEP, OZD, and TEHEED. F. 3 Mass spectra of impurities in the DEA feed Figures F.12 and F.13 show the mass spectra of MEA and TEA. F.4 Mass spectra of associated compounds Figures F.14 to F.16 show the mass spectra of BHG, DEEA, and MDEA. BHG or bis-(hydroxyethyl) glycine has been considered to be a degra-2 dation compound by some authors. However, no trace of BHG was found in the degraded DEA solutions and no mechanism seems feasible for its production under present experimental conditions. DEEA or diethylethanolamine is a tertiary amine with one hydrogen substituted with a hydroxyethyl group and the other two hydrogens substi tuted with ethyl groups. F.5 Summary Table F.l gives the molecular formula and the mass of the ions producing the major peaks in the mass spectra of each compound analysed. 100 1 -? i 4 80 -60 -56 40 -.20 -1 S6 112 122 143 157 174 207 rt - .1 Ii Hi | , .! i • > i 40 i 1 1 60 1 • • I :0 .,...,|.|.„,,....! • ,..,.| 1 0 0 1 2 0 ' i 1 40 160 1*0 i 1 i 1 •' 200 Figure F.l Mass spectrum of DEA ho 00 ioe so 60 42 100 14; 40 20 1 13 125 156 40 llll Hi 1 0 0 120 140 Figure F.2 Mass spectrum of BHEP 100 T so H 6L1 1 0 0 40 42 £0H 64 74 131 154 40 60 •I • I -ft 0 1 0 0 —i ' r"—• 1 1 r 120 140 Figure F.3 Mass spectrum of HEOD CO o llQO 30 60 40 20 0 J 1100 30 H 60 40 H 40 1 1 S sc. 1 0 0 ll li,,,, ,11111 11 11, I i u II , Jill i- .|HJ-M-i-130 I • I-J 1.1,1. 60 1 OO 156 •t" 1 74 1 8 199 207 £17 —, —, • 160 ISO V M 11 120 143 140 £4: 240 Figure F.4 Mass spectrum of THEED OO 100 80 H 60 i 40 A 20 H 44 l 74 56 40 -, T r-60 97 I I so 117 11. •. 11 129 157 149 i 1 ' 'I 1 OO 120 140 Figure F.5 Mass spectrum of BHEED ho CO Figure F.6 Mass spectrum of HEED ho CO CO Figure F.7 Mass spectrum of HEI oo Figure F.8 Mass spectrum of HEM K5 CO 100 ~t so 40 20 H I 50 60 • •nlllll.l I l Q'3 112 130 40 70 80 90 1 00 " i' " ' i' 1 10 " 'i • " M 1 120 1 30 Figure F.9 Mass spectrum of HEP CO ON 4 0 50 to ' 70 SO 90 100 1 10 Figure F.10 Mass spectrum of OZD ho CO Figure F.ll Mass spectrum of TEHEED ho Co 00 1 100-1 1 80 -60- 3 0 40 -20 -0 - I 1 • , 42 . ..hi . 61 , J 69 7 7 86 94 105 1 1 1 1 1 20 I ' i 40 1 60 1"' 1 1 80 100 Figure F.12 Mass spectrum of MEA 100 -j 80-60-40 -20 - 1 0 - • i • . • 1 • 1 11 40 I 1 i ' 80 I • 1 i ' 20 i 1 i 1 160 ii-200 i ' i 240 i ' r—• r 280 Figure F.13 Mass spectrum of TEA t-o CO VD Figure F.14 Mass spectrum of BHG 100 ~\ I 50 -n 40 20 i— 20 42 "t 4 0 102 -i——t-^ 1 00 1 1 Figure F.15 Mass spectrum of DEEA Figure F.16 Mass spectrum of MDEA r-o SO Table F.l Molecular formula and major peaks of mass spectra of compounds studied Compound M.W. Major Peaks Parent ion BHEED N,N-bis-(hydroxyethyl) ethylenediamine H0-C2H„ C2H4-OH \ / N-CzH^-N 148 74, 56, 44 / \ H H BHEP N,N-bis-(hydroxyethyl) piperazine HO-C2H4-N N-C^-OH 174 156,143,125,113, 100,98,70,56,42 BHG N,N-bis-(hydroxyethyl) glycine H0-C2Hu 0 II N-CH2-C-0H 163 118,74,56,45 H0-C2Ht, DEA Diethanolamine HO-C^ H0-C2Hi, N-H 105 74,56,45 ro Table F.l (cont.) Compound - M.W. Major Peaks Parent ion DEEA Diethylethanolamine C2Hb \ C2H5 H4-OH 117 86,58,42 / HEED N-(hydroxyethyl) ethylenediamine \ N / H H / -C^-N \ H 104 74,56,44 -* 0 HEI N-(hydroxyethyl) itnidazolidone / HO-C2H4-N 1 CH2 ''\ N-H 1 CH2 130 100,99,70,56,42 / HEM N-(hydroxyethyl) ethylenimine / HO-C2H.,-N \ CH, / \ CH2 87 56,42 / ho Table F.l (cont.) Compound M.W. Major Peaks Parent ion HEOD 3-(hydroxyethyl)-2-oxazolidone 0 II C / \ H0-C?H.,-N 0 I I CH2 CH2 131 101,100,56,42 / HEP N-(hydroxyethyl) piperazine C?H H0-C2H^-N N-H \ / C2HL, 130 112,99,88,70,56, / 42 MDEA Methyldiethanolamine H0-C2Hi, HO-C2H1, N-CH3 119 88,44 MEA Monoethanolamine N-C2H4-0H 61 61,30 Table F.l (cont.) Compound M.W. Major Peaks Parent ion 0 a OZD Oxazolidone 1 c / \ H-N 0 1 1 CH2 CH2 87 87,59,42 / TEA Triethanolamine HO-C2H4 \ N-C?Hi»-0H / H0-C2Hi, 149 118,56,31 -TEHEED N,N,N,N-tetra-(hydroxyethyl) ethylenediamine HO-C2Hu C2H>,-OH \ / N-C2H4-N / \ HO-C2H4 C2H4-OH 236 118,100,88,75, 56,45 -THEED N,N,N-tris-(hydroxyethyl) ethylenediamine HO-C2H1| C2H4-OH \ /. N-C2H4-N / \ HO-C2H4 H 192 143,118,100,88, 70,56,42 -Note: The underlined major peaks represent the ion with the relative abundance of 100%. Publications: 1. Kennard, M.L. and Meisen, A., "Aerosol Collection in Granular Beds" The 2nd World Filtration Congress, 1979, London, pages 229-238. 2. Kennard, M.L. and Meisen, A., "Degradation of Diethanolamine Solutions", paper presented at the Annual Meeting of the Canadian Gas Processors Association, 1979, Calgary. 3. Kennard, M.L. and Meisen, A., "Control DEA Degradation", Hydrocarbon Processing, April 1980, pages 103-106. 4. Meisen, A. and Kennard, M.L., "Practical Aspects of DEA Degradation Studies", paper presented at the 3rd Quarterly Meeting of the Canadian Gas Processors Association, Calgary, 1981. 5. Meisen, A. and Kennard, M.L., "DEA Degradation Mechanism", Hydrocarbon Processing, Oct. 1982, pages 105-108. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
United States 75 6
Russia 32 0
Republic of Korea 24 13
Canada 17 2
India 15 2
China 9 8
Iran 9 0
Italy 7 0
Germany 7 5
United Kingdom 7 0
Kuwait 6 0
South Africa 5 2
Ukraine 5 0
City Views Downloads
Unknown 139 21
Mountain View 8 0
Penza 6 0
Ashburn 6 0
Johannesburg 5 2
Shenzhen 5 8
Hyangyang 4 0
Arlington Heights 4 0
Stafford 3 0
Wilmington 2 0
Montreal 2 0
Seattle 2 0
Leeds 2 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0058802/manifest

Comment

Related Items