International Conference on Gas Hydrates (ICGH) (6th : 2008)


Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


59278-5627.pdf [ 138.65kB ]
JSON: 59278-1.0041102.json
JSON-LD: 59278-1.0041102-ld.json
RDF/XML (Pretty): 59278-1.0041102-rdf.xml
RDF/JSON: 59278-1.0041102-rdf.json
Turtle: 59278-1.0041102-turtle.txt
N-Triples: 59278-1.0041102-rdf-ntriples.txt
Original Record: 59278-1.0041102-source.json
Full Text

Full Text

Proceedings of the 6th International Conference on Gas Hydrates (ICGH 2008), Vancouver, British Columbia, CANADA, July 6-10, 2008.  GAS HYDRATE GEOHAZARDS IN SHALLOW SEDIMENTS AND THEIR IMPACT ON THE DESIGN OF SUBSEA SYSTEMS David Peters and Greg Hatton∗ Shell Global Solutions (U.S.) Inc. Westhollow Technology Center 3333 Highway 6 South Houston, TX 77082-3101, USA Ajay Mehta Shell Malaysia Exploration and Production Locked Bag No. 1, 98009 Miri, Sarawak, Malaysia  Chris Hadley Shell Exploration and Production Inc. 200 North Dairy Ashford, Houston, TX 77079, USA  ABSTRACT Gas hydrates in near-mudline subsea sediments present significant challenges in the production of underlying hydrocarbons, impacting wellbore integrity and placement of subsea equipment. As the fluids of an underlying reservoir flow to the mudline, heat carried by the fluids warms nearwell sediments and dissociates hydrates, which releases gas that can displace and fracture near well soil. This gas release may be calculated with numerical simulations that model heat and mass transfer in hydrate-bearing sediments. The nature and distribution of hydrates within the sediments, the melting behavior of the hydrates, the thermal and mechanical properties of these shallow sediments, and the amount of hydrates contained in the sediments are required for the model simulations. Such information can be costly to acquire and characterize with certainty for an offshore development. In this information environment, it is critical to understand what information, processes, and calculations are required in order to ensure safe, robust systems, that are not overly conservative, to produce the hydrocarbon reservoirs far below the hydrates. Keywords: gas hydrates, geohazard, wellbore, subsea system  INTRODUCTION The conditions exist in deepwater and arctic environments for gas hydrate stability and gas hydrates may occur in the near seafloor soils. A number of models have been developed to evaluate strategies to produce these hydrates as a potential source of methane gas. In addition, ∗  studies have been performed on drilling through hydrate-bearing sediments. However, only recently has the production of hot hydrocarbons via wells through hydrate-bearing layers been analyzed. Such production heats and dissociates near-well hydrates and releases gas. Hazards associated with this gas release may include displacement and fracturing of near-well soil and  Corresponding author: Phone: +1 281 544 7420 Fax +1 281 544 8826 E-mail:  associated failure of the integrity of subsea components—such as the wells and well jumpers. In order to safely produce hydrocarbon reservoirs far beneath near-mudline hydrates, the risks associated with wells that pass through hydratebearing sediments must be understood and managed. These wells may produce for 30 years, or longer, and raise the temperature of nearmudline sediments above the hydrate dissociation temperature for hundreds of meters from the well. This can result in the release of a large amount of gas that can impact the subsea system in several ways. The near-well formation (soil, hydrate, etc.) may fracture due to the volume change caused by dissociation of hydrates into water and gas. Volume changes will lead to displacement of the formation supporting the well and may cause casing integrity issues. Collet and Dallimore [1] have documented arctic drilling and production hazards attributed to gas hydrates. Movement of the formation further from the well can lead to the displacement of manifold relative to the well, which can lead to well jumper integrity issues. The impact of hydrate melting on the integrity of subsea systems may be quantified with an understanding of the consequences of: 1) Heating the hydrate-bearing layers on the material around the well. 2) Changes in the material around the well on the subsea systems. The analyses for items 1) and 2) are coupled but generally performed separately. This paper presents the results of generic simulations performed by Shell to better understand the key input parameters and their impact on the overall system integrity. HYDRATE SIMULATIONS In order to assess the impact of melting hydrates on a subsea system, three general types of models are required. The first model simulates the heat flow and associated hydrate dissociation around the well. The second model simulates the geomechanical response of the formation to hydrate dissociation. The third type of model is a structural model of the well and well jumper to verify the integrity of these components based on the results from the first two models. The focus of this paper is on the first two models, for which  predictions generated with the ToughFx/Hydrate code [2] are shown. (An in depth discussion of integrity analysis is beyond the scope of this paper.) While these two types of models have been available separately for a few years, application and coupling of these models to a particular development requires significant geomechanical and hydrate expertise. Only recently have groups begun to develop coupled models [3,4]. Assessment of Hydrate Models Two important aspects to be considered in modeling the melting of hydrate in sediment are the thermal response of the system due to heating and the transport of released gas. Most models account for both of these aspects since they were developed to analyze different systems for producing the gas released by dissociating hydrates in the sediment, although the simplifying assumptions of the different models vary considerably. The thermal model simulates the heating of the material around the well by the fluids produced from the underlying reservoir. For this analysis, a good understanding of the thermal characteristics of 1) the well and associated casing, and 2) the surrounding material/sediments is necessary. Fortunately, the thermal response of the system is straightforward, as heat transfer in these types of systems is well understood. The transport of gas and water released by hydrate dissociation is a function of a number of the formation properties (including permeability, in situ stress state, shear strength and pore pressures) and is much more difficult than the thermal response to accurately represent in a model. There is still much debate as to fluid transport in the formation once hydrates dissociate (and pore pressure increase over that required to fracture the formation). Depending on the field and the fielddevelopment layout, the geomechanical mechanisms can have a large impact on transport. Fortunately, the risk due to melting hydrates can be assessed with limited geomechanical data by analyzing limiting cases. Two limiting cases useful for this type of analysis are 1) retain all of the gas released by hydrate dissociation in situ as could occur if the horizontal fractures are formed, and 2) release the gas from the formation as could  Assessment of Modeling Inputs The parameters that have the largest impact on the amount of hydrate melted in the formation are the hydrate saturation and those parameters related to the rate of heat transfer in the system. The amount of hydrate present in the formation to a large extent controls how much gas is released due to melting hydrate. However, getting an accurate assessment of the amount of hydrate in the near-seafloor formation is not a trivial task. In typical field development scenarios, early seismic and petrophysical data acquisition is directed to the deeper hydrocarbon reservoirs and the data sets may be inadequate for characterizing the near seafloor sediments. As a result, interpretation of the volume of hydrate is difficult to assess. Limited experience to date indicates that hydrate saturations evaluated from such data sets can be very inaccurate. A dedicated data acquisition program, such as that described by Trehu et al. [5] is needed to provide an accurate the assessment of gas hydrate saturations. In addition to hydrate saturation, the distribution of hydrates within the formation is an important variable. Evaluation of hydrate saturations at discrete borehole locations does not provide an understanding of the range over which any sampled hydrate deposits extend—i.e., whether the hydrate saturation is radially uniform for several centimeters, or several hundred meters, from the measurement site. A geologic model of a site should therefore be used to target borehole locations and support extrapolation of such data within a field development area.  This raises an important aspect related to the value of information. Numerous methods (with varying costs) have been used to measure hydrate saturation [5]. An assessment of hydrate saturation can be obtained through interpretation of petrophysical logs. The accuracy of this estimate is dependent on the formation properties and quality of the logs. However, depending on the formation properties, a better assessment of saturation may be obtained by retrieving pressurized cores of the hydrate-bearing soil. This allows a more direct measurement of the hydrate saturation at discrete core depths and provides a means to calibrate interpretations based on petrophysical logs. The use of pressurized cores to provide information on the gas composition of the hydrates at different locations is important to assess the ultimate depth of hydrates and the hydrate subcooling, as is illustrated in Figure 1. For example, if the gas composition changes as a function of depth, it is possible to get an increase in hydrate subcooling with decreasing depth. The hydrate subcooling is an important parameter in that it determines the temperature to which the sediment must be warmed before hydrates can melt. 10 site1 site2 site3  9.5 9 8.5 8 7.5 7 Hydrate Subcooling [C]  occur if vertical fractures intersect the seafloor. Quantifying the uncertainty of the analysis results for these two limiting cases is considerably easier than for other cases—which require more detailed knowledge of the geomechanical properties of the hydrate bearing sediments and associated detailed transport modeling. In the current study, the results of the limiting-case analyses and the welljumper-manifold integrity analyses of several field-development scenarios clearly bounded the geohazard risk and provided adequate information for the development scenario selection process. Thus, the limiting-case analysis, can be more readily used to screen various field development scenarios.  6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 0  50  100  150  200  250  300  Depth Below Mudline [m]  Figure 1 Hydrate subcooling at each drill center versus depth In order to estimate a reasonable average over the area of interest, determination of the hydrate saturation at several locations within the area of interest is recommended. In addition, if there is reason to believe that the hydrate forming gas is not pure methane, obtaining pressurized core samples is recommended so that gas chromatography may be used to determine the composition of gas released from the pressurized core.  The rate of heat transfer in the system is a function of the well design parameters, well flow rates, and hydrate bearing formation parameters. Although the models are sufficiently accurate to model heat transfer in such systems, the parameters used in the models have considerably less certainty. During production, heat is transferred to the formation from the hydrocarbons flowing in the well. The heat flows from the inner production tubing, through the various casing and annulus layers of the well, and into the formation. The design of the wellbore and all of the thermal properties of the various layers are required for this calculation. Care needs to be taken to accurately model the overall rate of heat transfer through the well and include possible effects due to convection cells in the fluid filled annuli. The thermal properties of the sediment are less well understood when hydrate are present. In particular, the thermal conductivity and heat capacity of hydrate bearing sediment and how these properties change as the hydrate melts is still not well understood. In addition to the thermal properties of the sediment, it is also important to understand the temperature and pressure increase with depth as these determine the depth to which hydrates remain stable and how much each depth must be heated in order to melt hydrate. As part of the data gathering process during drilling, the temperature and pressure should be measured. In addition, information on the thermal conductivity of the formation may be obtained by laboratory testing of core samples and through correlation with other formation properties (water content and density).  heat flux (from the well) going into the hydrate bearing sediment. The heat entering the soil either warms the soil or melts hydrate. The latent heat of hydrate dissociation is high relative to the heat capacity of the hydrate bearing sediment, which means that, for moderate and high hydrate concentrations, most of the heat melts the hydrate as opposed to warming the soil. As a rough approximation, it can be assumed that the amount of hydrate melted is roughly proportional to the heat flux. As a result, it is important to get a good estimate of the heat flux as it can be directly related to the amount hydrate melted and gas released. The heat flux changes with time due to the increasing temperature of near-well soil with time. As shown in Figure 2, initially the heat flux from the well is high due to the low temperature of the near-well soil relative to the well temperature. With time, the heat flux decreases as the soil temperature increases. Since the formation characterization varies with the formation stratigraphy, the heat flux is calculated for each strata (layer). Typically, a complete representation of the stratigraphy by depth is used for single-well simulations so that the effects of layer variation and an initially high heat flux are captured. The multiple well cases are simplified by use of an average heat flux for the duration of the simulation. There are a number of different ways to define the average heat flux, but typically the “average” heat flux is chosen such that the mass of hydrate melted at the end of the average-heat-flux simulation is the same as it would be were the actual heat flux used. 600  500  Simulation Examples  Perhaps the most important parameter for estimating the rate of hydrate dissociation is the  Heat Flux [W/m]  400  In order to assess the risks to subsea system design as a result of hydrate melting, two types of simulations should be performed. Well integrity can be evaluated with single-well simulations, to examine the impact of near-well melting of hydrates. Jumper integrity can be evaluated with multiple-well simulations in order to understand the relative displacement of the jumper hubs at the well and the central manifold.  Actual Heat Flux Single Well Cases  300  Average Heat Flux Multiple Well Cases 200  100  0 0  1000  2000  3000  4000  5000  6000  7000  8000  9000  10000  Elapsed Time [days]  Figure 2 Example of the Heat Flux from the well into the surrounding soil at a given depth  220 site1 210  site2 site3  200  Average Heat Flux [W/m]  190 180  170 160  230 Cp=664 J/kgC Cp=920 J/kgC 220  210 Average Heat Flux [W/m]  The thermal conductivity, κ, of the hydrate bearing sediment is also an important parameter for each formation layer as the heat flux is roughly proportional to κ times the temperature difference. Figure 3 shows the correlation of the thermal conductivity and the heat flux. Note that each point on the figure represents a different layer (depth in the system). As would be expected, the heat flux is roughly linearly correlated with the thermal conductivity. Two of the simulations performed are at roughly the same pressure/temperature conditions, so the scatter for these locations is simply due to the variation of temperature with depth. For the third simulation, the heat flux is shifted towards lower values at a given thermal conductivity due to very different temperature conditions. Also note that heat flux and thermal conductivity each can vary over a wide range for a system with similar characteristics. If the specified thermal conductivity for the simulations is too high, then the calculated reservoir temperatures are too low and the calculated heat fluxes are too high.  200  190  180  170  160  150 1.60  1.70  1.80  1.90  2.00  2.10  2.20  2.30  2.40  2.50  Thermal Conductivity [W/m/C]  Figure 4 Heat Capacity versus Average Heat Flux The primary output from the thermal simulations is the amount of gas released as a function of time. Figure 5 shows a plot of the total volume of gas released due to hydrate melting for a broad range of formation parameters. The volume of gas is a strong function of the hydrate saturation. All of the heat that enters into the soil is used either to heat the formation or to melt hydrate. In cases with low hydrate saturation, a greater percentage of the total heat entering the system is used to heat the soil, whereas with the high hydrate saturation, a greater percentage is used to melt hydrate. Scatter around the best-fit lines is due to the variation in one of the many other simulation parameters.  150 45000  140 40000  120 1.25  1.45  1.65  1.85  2.05  2.25  2.45  Thermal Conductivity [W/m/C]  Figure 3 Thermal Conductivity versus Average Heat Flux  Gas Evolved Due to Melting Hydrate after 30 years [kg/m]  130  35000  30000  25000  20000  15000  10000  The second parameter that is important in characterizing the heat transfer through the hydrate bearing sediment is the heat capacity. In our simulations values of the heat capacity of the formation between 920 J/kgK and 664 J/kgK are assumed to bound the range of the expected values [6]. Note that these are not the actual heat capacity of the soil matrix, but rather these values are chosen to match the heat capacity of the composite hydrate bearing sediment. The correlation between heat flux and thermal conductivity for the simulated formation is shown in Figure 4.  5000  0 0  5  10  15  20  25  30  35  40  45  Hydrate Saturation [% of total volume] Multiple Well Case  Single Well Case  Figure 5 Amount of Gas Released as a Function of the Initial Hydrate Saturation for Single Well and Multiple Well Simulations Figure 5 also shows the effect of multiple wells on the amount of gas released. For these cases, the total volume of gas released is determined and then divided by the number of wells to get the  In Figure 5, the one-well and multiple-well results show similar trends but provide differing results. This leads to the important question of what level of detail is needed in the simulation. With all of these simulations, there is a trade-off between accuracy of the predictions and computational time. (For example, the information obtained from the resistivity logs is sufficient to define the hydrate saturation for many more layers than the simulator can process in a week.) Keeping in mind that the primary output variable is the volume of gas released, we ran a sensitivity analysis and determined that it is possible to significantly reduce the level of detail needed in the simulation and still calculate the volume of gas released to the desired accuracy. However, care must be taken to provide appropriate average hydrate saturations to calculate the released gas volume. Figure 6 shows a typical representation of the hydrate saturation in the system for a single well simulation. The melting hydrate interface is clearly shown in the figure and it is not uniform. The hydrate interface for layers with lower hydrate saturations is further from the wellbore than the interface for layers with higher hydrate saturation. But as shown in Figure 5, this does not necessarily mean that more gas is evolved. SYSTEM DESIGN The intent of the hydrate simulations is to estimate the potential impacts of hydrate melting on the integrity of the well and the jumper between the wellhead and the manifold. In the case of well integrity, the well must be designed to withstand the strains transferred into the well due to the displacement of the formation adjacent to the well during the initial years of production. As discussed previously, the volume of gas released as a function of time and position are the most important outputs from the hydrate-  reservoir simulation in regard to analyzing well integrity. 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12  -50  Depth below mudline [m]  volume of gas released per well. In all cases the volume of gas released with multiple wells is less, on a per well basis. This is largely due to the overlapping zones of heating for multiple well cases. Once the heating zones begin to overlap and the hydrate is melted, the addition of more heat increases the temperature of overlap area rather than dissociate hydrates.  -100  -150  -200  10  20  30  40  50  60  70  80  90  Radial distance from well [m]  Figure 6 Hydrates Saturation from a Single Well Simulation After 30 years of Production The same is true in understanding the effects of hydrate melting on jumper integrity. Here the issue is the relative displacement of the jumper hubs at the wellhead and the manifold, i.e., at the two ends of the jumper. If some reasonable estimate of the radial extent of hydrate melting is determined, for cases with multiple wells, the jumper can be designed to withstand the strains caused by the hydrate induced well and manifold motions. To produce a hydrocarbon reservoir under hydrate bearing sediments the field layout may: 1) Produce directly up through the hydratebearing zone above the reservoir 2) Avoid producing through the hydrate-bearing zone (by offsetting the wellheads and manifolds to an area without hydrates) 3) Produce through a low-concentration hydrate-bearing zone (by offsetting the wellheads and manifolds to an area over lower hydrate concentrations) In concept, the second field-layout option is straightforward. However, while relocating equipment is possible, the associated cost—e.g., longer and more deviated wells to reach the hydrocarbon reservoir of interest—may be very large. Usually, the first and third options, if they can be designed to be safe, are less costly than the second option. These alternatives require well and jumper designs that are sufficiently robust to  handle the respective hydrate issues. In all cases, good understandings of the formation properties—including hydrate distributions and gas release—and of the well strength are required, so that the system can be adequately designed to manage the risks without being excessively expensive. CONCLUSIONS Producing hydrocarbons through hydrate bearing, near-mudline sediments, heats and dissociates hydrates. This process releases gas causing a volume change in the formation that can displace the formation and strain wells, and well jumpers, potentially leading to a loss of structural integrity. Managing the geohazard risks for wells and well jumpers requires: 1. A sound understanding of the formation response to transporting hot fluids through the hydrate bearing layers. 2. Well and jumper structural analysis that incorporates the results from item 1. With these, the gas hydrate risks to well and jumper integrity can be quantified and managed for different field development scenarios. The most important output from the hydratebearing-layers simulations that goes into the integrity analysis is the quantity of gas released as hydrates dissociate. Thus, understanding the relation between dissociated gas quantities and the simulator inputs is critical. To better understand this, sensitivity studies were performed. The results of these sensitivity studies establish the value of information of input parameters. That is, the value of knowing the nature and distribution of hydrates within the sediments, the melting behavior of the hydrates, the thermal and mechanical properties of the formation and the amount of hydrates contained in the sediments. For reservoirs with severe geohazard risks, the hydrate concentration is a critical reservoir parameter. Understanding the key parameters that determine the stability of the material around the well and the relation between this and well system integrity provides a project team with a good understanding of which parameters need to be measured and which ones can be estimated. With a solid understanding of the impact of key parameters and  the tools necessary to determine well system integrity, a project team can effectively manage risk and select an appropriate field development plan. REFERENCES [1] Collett, T.S., and Dallimore, S.R., Detailed analysis of gas hydrate induced drilling and production hazards, In the Proceedings of the Fourth International Conference on Gas Hydrates, April 19-23, 2002, Yokahama, Japan, 8 p. [2] Moridis, G., Kowalsky, M., and Pruess, K., ToughFx/Hydrate v1.0.1 user’s manual: a code for the simulation of system behavior in hydratebearing geologic media. 2005, LBNL/PUB 3185. [3] Rutqvist, J., and Moridis, G., Numerical studies on the geomechanical stability of hydratebearing sediments. 2007 Offshore Technology Conference, OTC 18860. [4] Freij-Ayoub, R., et al., A wellbore stability model for hydrate bearing sediments. Journal of Petroleum Science and Engineering, 2007:57:209220. [5] Trehu, A. M., et al. Three-dimensional distribution of gas hydrate beneath southern Hydrate Ridge: Constraints from ODP Leg 20”. Earth and Planetary Science Letters, 2004, 222:845-862 [6] Ochsner, T.E., Horton, R., and Ren, T., A new perspective on soil thermal properties. Soil Science Society Journal, 2001:65:1641-1647.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items