Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Some physical characteristics and exposure loss estimates of pea vine silage in a horizontal silo Campbell, William Earl 1959

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

Item Metadata


831-UBC_1959_A4 C2 S6.pdf [ 8.1MB ]
JSON: 831-1.0059073.json
JSON-LD: 831-1.0059073-ld.json
RDF/XML (Pretty): 831-1.0059073-rdf.xml
RDF/JSON: 831-1.0059073-rdf.json
Turtle: 831-1.0059073-turtle.txt
N-Triples: 831-1.0059073-rdf-ntriples.txt
Original Record: 831-1.0059073-source.json
Full Text

Full Text

SOME PHYSICAL CHARACTERISTICS AMD EXPOSURE LOSS ESTIMATES OP PEA VINE SILAGE IN. A HORIZONTAL SILO by WILLIAM EARL CAMPBELL B.S.A.(Hons.) University of British Columbia, 1953 A THESIS SUBMITTED II PARTIAL FULFILMENT' OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN AGRICULTURE in the Department of AGRICULTURAL MECHANICS We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April, 1959 i ABSTRACT Reports of large losses from exposed horizontal silos under high f a l l and winter r a i n f a l l were investigated by f i e l d observation on a f u l l scale test s i l o . Density, dry matter, permeability and some proximate analyses of the silage are tabulated. Observed dry matter losses in order of magnitude are fermentation losses, sub surface losses, surface losses, and losses i n runoff and effluent liquids. The value of these losses was computed. The best method of reducing the storage losses was a plastic sheet l a i d over the silage immediately after f i l l i n g and packing. In presenting t h i s thesis i n p a r t i a l fulfilment of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y 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 representatives. It i s understood that copying or publication of this thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of A g r i c u l t u r a l Mechanics The University of B r i t i s h Columbia, Vancouver 8, Canada. Date May 4, 1959.  i i TABLE OF CONTESTS P a g e Introduction , 1 Literature Heview 1 Objectives of the Investigation 3 Materials and Methods Materials 3 Methods 4 Observations 11 Discussion 27 Conclusion 47 Literature Cited 49 Appendix 52 i i i LIST OF TABLES Table Page 1. Summary of r a i n f a l l by months, 1957-58 and 1959 11 2. Floor effluent summary by months, 1957-58 13 3. Floor effluent summary by months, 1958-59 14 4. "Wall" effluent summary by months, 1958-59 15 5. Runoff summary by months, 1958-59 16 6. Protein and ash in dry matter from floor e f f l u -ent and runoff, 1958-59 16 7. Pea vine and silage sample summary. 17 8. Proximate composition of 1956-57 silage samples 18 9. Proximate composition of 1958-59 silage samples 18 10. Sub surface layer thickness during the storage period. 22 11. Permeability of core samples, March 6, 1959 23 12. Calculated storage losses and yields, 1958-59 25 13. Percent dry matter surface loss, and yield for grass silage exposed and covered by a plastic sheet, January 8, 1959. 26 14. Rainfall and total effluent collected by months, 1958-59 32 15. Calculated liquid balance September 1958 - March 1959 33 16. Summary of dry matter losses i n effluent liquids, 1958-59 35 17. Calculated dry matter balance September 1958 to March 1959 42 iv LIST OF FIGURES Figure Page 1. Plan and side elevation of liquid collection system, 1957-58 4 2. Layout of test area, 1957-58 5 3. Silage sampling: tool 7 4. Final sampling tool 7 5. Piezometer, showing construction 8 6. Piezometer arrangement 9 7. Collection pan on vertical face of silage 9 8. Corer for permeability tests 10 9. Permeability test arrangement 10 10. Method of f i l l i n g the si l o 12 11. Silage after tractor packing August, 1957 12 12. Silage after tractor packing August, 1958 13 13. Surface conditions November, 1957 19 14. Surface conditions February, 1959 19 15. Surface conditions December, 1957 20 16. Appearance of exposed face, February 1958 24 17. Rainfall and floor effluent collected by months, 1957- 58 30 18. Rainfall and effluent collected by months,1958-59 31 19. Dry matter losses i n effluent, by months, 1957-58 and 1958-59 34 20. Dry matter distribution and yield, 1957-58 and 1958- 59 44 V APPENDIX Page A. Rainfall, 1957-58 season 53 B. Rainfall, 1958-59 season 55 C. Effluent flow from floor area, 1957-58 57 D. Effluent flow, 1958-59 58 E. Piezometer performance 63 F. Permeability equation 64 v i The author wishes to acknowledge the assistance given by the follow-ing individuals: Dr. J.C. Berry, Dr. V.C. Brink, Prof. T.L. Coulthard, and Prof. B.L. Watson. Be also wishes to acknowledge the cooperation of the Department of Agricultural Mechanics, Division of Animal Science, and the Division of Plant Science. -1-INTROIXJCTION A number of enquiries by farmers in the Fraser Valley sug-gested that high storage losses, particularly in drainage l i -quids, occurred in horizontal silos exposed to rain. Since l i t t l e published work; was available on dry matter losses in effluent from horizontal silos exposed to high r a i n f a l l , a stu-dy of storage losses was undertaken. This work was not intended as a chemical or nutritional study of silage or silage fermentation, but rather as an e s t i -mate of storage losses and physical characteristics of silage during the storage period. It was intended to provide data on storage losses and suggest methods of keeping storage losses to a minimum. LITERATURE REVIEW A relatively large number of references were found on the general topic of silage effluent losses, but nearly a l l referr-ed only to effluent from tower s i l o s . Only a few mentioned the approximate dry matter losses in the effluent. Barnett (3) gives data on the composition of effluent from tower silos and separates compression losses from the fermentation losses. He also emphasizes the need for free drainage from tower sil o s , to prevent the formation of unpalatable silage, even though slight-ly higher dry matter losses were observed as a result of free drainage. S i l l and Sears (17) reported that a roof reduced the leaching losses in tower s i l o s . Watson (19) gave some of the factors contributing to ef-fluent loss in tower silos, including the water content and the silage pressures as variables, and noted that no leaching loss-es occurred unti l after rain had fallen on the s i l o . Dry mat-ter losses in the effluent of a roofed tower sil o were given as 1.5%, and an estimate of losses in effluent for roofed tow-er silos generally was given as less than 5% of dry matter i n -put. Cooper et al (7) reported storage losses in small experi-mental silos as 17.5% of input, and fermentation losses as 14.5% (average of 2 years). The use of plastic covers reduc-ed these losses to 4.5% and 5*5% respectively. Total silage losses of 32.4% of input material were reported, but no expos-ure conditions were given. Le Clerc (9) gave dry matter losses in effluent from tower silos as 47-82 pounds per one hundred gallons effluent with the dry matter analysing 17-28% protein, 13-23% ash and amount-ing to an appreciable loss. Blish (5) however, stated that effluent losses for sunflower'silage in tower silos were neg-l i g i b l e , and noted that free drainage was desirable to prevent waterlogging and formation of sour silage. Waterson (18) reports that rainwater soaking through the silage in a horizontal s i l o may cause secondary fermentation under the "slimed over" seal which forms after r o l l i n g and packing of the silage surface. Nash (18) states that a roof i s required under Scottish conditions, and large losses have occurred in unprotected silos in wet years. He reports that maintenance of dry matter content i s important to maintain milk yield, stating that milk yields usually f a l l when low dry matter silage replaces high dry matter silage in the ration. He also reports 20% dry matter silage as being less than op-timum dry matter. Density of silage has been reported i n some detail . Otis and Pomroy (14) reported very large variations in silage density a-cross the diameter of a tower s i l o . Esmay et al (8) reported that average lateral pressure in a horizontal silo was relatively con-stant with increasing depth, below two feet. OBJECTIVES OF THE INVESTIGATION The objective of this investigation was to measure the magni-tude and distribution of losses due to seasonal r a i n f a l l on an exposed horizontal s i l o . The volume and dry matter content of the l i q u i d running off the surface, passing down along the walls and draining from the floor of the silo was measured. The den*-sity and permeability of the silage was investigated, as these characteristics could affect the leaching and surface losses. The changes in total dry matter and dry matter percent of the silage over the storage period and the dry matter losses i n i n -edible or unpalatable silage was measured. The cost of the preventable storage losses was then used to determine what meth-ods of protecting the surface were feasible. MATERIALS AND METHODS Materials This investigation was conducted on pea vine silage since pea vine was a low cost, readily available material which had shown high losses in horizontal s i l o s . Because i t was a rela-tively coarse, unchopped material, i t was f e l t that i t would probably be quite susceptible to both surface losses and leach-ing. -4-Methods It was assumed that the silage would behave very much like a s o i l , where percolation i s primarily vertical, and that hor-izontal flow i s negligible except where Impervious layers i n -terfered with normal vertical flow. It was assumed that the normal flow pattern in silage would be runoff over silage sur-face toward the "toe" of the s i l o , and vertical percolation of a relatively small part of the r a i n f a l l to the sil o floor. I f the drainage provided under the silage was not adequate to allow removal of the vertical percolation, waterlogging could occur. In order to measure the volume and source of liquid reach-ing the floor of the s i l o , a test area fifteen feet long was prepared near the ramp end of a fifteen foot wide s i l o ( F i g . l ) . 15' r Guard Length v Guard Drain F l o o r Area "Toe" of s i l o >Wall D r a i n To Tanks s 1 5 » — P e r c o l a t i o n Test Length Pea Gravel' ' - - F l o o r Drains •—,' Guard Length .Guard Drain. Ramp P l a s t i c Sheet F i g . l . Plan and side elevation of liquid collection system, 1957-58. A collection system designed to allow separate collection of liquids from different areas was installed. Any effluent c o l l -ected from the floor area, fifteen feet long and fourteen feet wide from wall drain to wall drain, was designated "floor" ef-fluent. Any liquid collected by the pipes laid parallel to the walls beside the floor area was designated "wall" effluent. Drain pipes were also installed i n the seven foot long guard area at the ends of the test area to reduce "edge effects". The pea gravel and sand layer was intended to simulate a well drained s i l o . Pig. 2 i s a picture of the site before the collection pipes were completely covered. Fig. 2. Layout of test area, 1957-58. A similar system was prepared in 1958, and a plastic sheet long enough to extend the f u l l length of the s i l o was i n s t a l l -ed with a drain at the "toe" end of the s i l o . Any liquid c o l l -ected from this "toe" drain, combined with the liquid from the guard area drains, was designated "runoff", since most of the liquid came from surface runoff over the silage. F i l l i n g of the s i l o was done carefully to prevent damage to the drainage collection apparatus. Rainfall for the area during the period from f i l l i n g to the end of tests was obtained from the weather records of the Van-couver (U.B.C.) Station, located within one eighth of a mile from the s i l o . Samples of the fresh pea vine, effluent liquids and silage were taken. Pea vine "grab" samples were taken as the s i l o was being f i l l e d and oven dried at 100°G for twenty-four hours or to constant weight. Effluent samples were taken at the start of effluent flow, and as required to follow changes in percent dry matter in the effluent. The effluent samples taken prior to October 1, 1958, were oven dried at 70°G. Effluent samples taken after October 1st were dried under vacuum to reduce the charring noted in oven dried samples. Protein and ash analys-es by A.O.A.C. methods were done on "runoff" and "floor" ef-fluent collected in October and at the end of the exposure period. Silage samples were taken near September 1st and January 31st, and as deemed neoessary from surface or feeding require-ments. Since the original volume and weight of the samples was known, the density was easily calculated. Silage samples were oven dried to constant weight at 70°C. Proximate analys-es of the September 1958 and March 1959, and some 1956-57 samples were done by A.O.A.C. methods. Successive silage samples were taken from selected adja-cent undisturbed parts of the percolation test area of the Fig, 3. Silage sampling tool. Fig. 4. Final sampling tool, A relatively undisturbed solid core, similar to the core in the l e f t in Fig. 3 can be obtained from silage up to six feet in depth with the longer sampler used after December 1957 as shown in Fig. 4. The hole l e f t after sampling was packed f u l l of silage from an adjacent s i l o , to maintain the original con-ditions of density, moisture movement, and aeration during the storage period. For the purpose of this study, to distinguish fermentation and storage losses, the storage period was defined ar b i t r a r i l y as the period between one month from f i l l i n g to the end of ob-servations. The silage losses observed were separated into "spoil layer" losses and "sub surface layer" losses. The sur-face or "spoil layer" was defined as the material above the originally properly fermented silage, and included a partly decomposed surface layer, a slimy rotted layer which i s the actual seal layer, and a very thin partially decomposed lay-er below this seal layer. The "sub surface" layer was that layer of material under the "spoil layer" which had been prop-erly fermented, but had subsequently been contaminated by sur-face products or had undergone other changes making i t unpa-latable to cattle. LEAD WIRE ANCHOR BLOCK CONTACT WIRE PLUG Fig. 5. Piezometer, showing construction. -9-Several other experimental techniques were tried. Since ob-servations in 1957-58 suggested that the nsub surface" layer might be saturated, electric piezometers which could indicate saturation were installed at different depths in the silage (Fig. 5 and 6). \> 1 0 ' -y L fm Floor 'j Area * 10- H O 1 < _"Toe M'of the Ground ! Rod w h 3^ | silage "* Ditch and Dam for the "Runoff" System .4 lie ads brought to the wall - Subscripts indicate depth in feet Fig. 6. Piezometer arrangement. Fig. 7. Collection pan on vertical face of silage. -13-Another technique was used to measure horizontal movement of water in the "sub surface" layer at the exposed face of the silage. A collection pan one foot square was fixed into the vertical face so that any effluent flow from the silage could he measured (Pig. 7) . Soil permeability techniques were used to measure the perm-eability of core samples taken horizontally and vertically from adjacent positions in the silage (Pig. 8 and 9). COREP. -11-Some other observations were made during the investigation. The unpacked depth after f i l l i n g , the packed depth, settling rates, surface conditions, "spoil" and "sub surface" layer depths and any other relevant information was recorded. Some observa-tions and one pair of samples were drawn from an adjacent s i l o , one half of which was protected by a sawdust covered plastic sheet, to estimate the effect of the surface protection. OBSERVATIONS The r a i n f a l l , by months, from f i l l i n g to the end of observa-tions, i s summarized i n Table 1, from Tables A and B in the appendix. Month Rainfall In Inches 1957-58 1958-59 August 0.29* 1.86 September 1.16 2.16 October 3.24 4.26 November 3.67 7.29 December 6.72 8.11 January 11.25 6.78 February 5.77 5.74 Total 32.10 36.20 * records started August 20th. Table 1. Summary of r a i n f a l l by months from August to February inclusive, 1957-58 and 1958-59. In both 1957 and 1958 the si l o was f i l l e d and packed by a tractor equipped with a front end loader, which carried the mat-er i a l into the si l o from truckloads dumped nearby (Fig. 10). In 1957, the pea vine was very wet, as low as 15% dry matter, and -12-some showers occurred during f i l l i n g . During packing, the com-paction pressures under the wheels of the packing tractor were high enough to express moisture from the wet pea vine. The pea vine packed solidly and showed l i t t l e "spring back" (Fig. 11). In 1958 the pea vine was considerably drier, averaging 21.7% dry matter, and did not pack well. The surface remained rela-tively loose even after packing (Fig. 12). Showers just after packing did not improve the surface compaction. Fig. 10. Method of f i l l i n g the s i l o . Fig. 11. Silage after tractor packing August, 1957. -13-Fig. 12. Silage after tractor packing August, 1958. The very wet pea vine ensiled in 1957 started losing e f f l u -ent almost immediately after f i l l i n g (Table 2). The "floor" effluent amounted to nearly one gallon per day the f i r s t ten days, then decreased to about two gallons per week for the next two weeks. After September 5, "floor" effluent losses were very small . The percent dry matter in the 1957 "floor" effluent rose slightly over the storage period. A summary of "floor" effluent observations i s presented i n Table 2, from Table C i n the appen-dix. Month Volume Gallons Percent Dry Matter Dry Matter in Pounds August September October November December January Total 14.69 .75 .75 1.60 2.30 50 20.59 gal. * estimated 3.72 4.05 4.35* 4.65* 4.95* 5.27 5.5 .3 1.1 .2 8.1 lbs, Table 2. Floor effluent summary by months, 1957-58. The percent dry matter estimated for October, November, and Dee-ember were based on linear interpolation between September and January observed percent dry matter i n the floor effluent. The "wall" effluent was less than two gallons for the whole period from August 1st to January 31st and the "guard" effluent volume, though not measured, was observed to be very small. In 1958 effluent flow from the "floor" area was negligible u n t i l September 8. From September 8 to September 28, a consid-erable effluent flow was measured. After September 28, only a very small volume of effluent was collected as shown i n Table 3, summarized from Table D i n the appendix. Month Volume Percent Dry Matter Gallons Dry Matter in Pounds August 0.1 10.7* 0.1 September 29.33 10.7 30.9 October 4.5 8.06 .36 November .98 6.63 .06 December .84 5.1* .04 January .62 3.7* .02 February .46 2.2 .01 Total 36.83 gal. 31.49 lbs. * estimated Table 3. Floor effluent summary by months, 1958-59. A considerable volume of wall effluent was collected in 1958, but the distribution was very uneven (Table 4). The con-necting pipe to the measuring tank appeared to be plugging at irregular intervals, and efforts to keep i t operating did not make any difference to effluent flow. A sudden very high rate of flow was observed between October 13 and October 18, when 119 gallons of effluent was collected i n five days. Another period of high flow ra tes occurred November 11 to 16th, after -15-which flow rates became almost uniform, with no sudden changes in rate (Table 4).. Month Volume Gallons August September October November December January February Total 1.40 3.20 121.0 57.85 12.0 32.5 73.2 304.15 gal. Percent Dry Matter 10.0 6.06 Use 6.06 as average Dry Matter in Pounds 1.4 3.2 73.5 35.2 7.3 19.7 44-4 184.7 lbs. Table 4. "Wall" effluent summary by months, 1958-59. The weight of dry matter i n the effluent was calculated using a value of 6.06$, which was observed during the period of maxi-mum flow. In 1958 "runoff" was collected from the "toe" of the s i l o . This liquid was collected from a total exposed area forty feet long by fifteen feet wide. The volume of liquid was high. A considerable amount of "runoff" was lost by overflow from the "toe" area i f r a i n f a l l exceeded an average of 0.3 inches per day for several days, and some liquid was lost from the c o l l -ection tanks overflowing. Whenever possible, conservative est-imates of the losses were made. The i n i t i a l high dry matter percent in the "runoff", 10$ on September 1, dropped to 1.39$ October 13 and 0.53$ by February 4th. The percent dry matter for lovember, December, and Janu-ary were estimated by interpolation between the October 13th and February 4th observed percent dry matter i n the "runoff". This gives a dry matter loss about twenty-one pounds greater -16-than the product of observed "runoff" times average percent dry matter for the period October 13 to February 4 inclus-ive. The observed dry matter losses were highest i n Aug-ust and September, and were appreciable every month. Month Volume of Percent Dry Matter "Runoff" Dry Matter in Pounds August 154.6 10.0 154.0 September 237.5 5.7* 135.0 October 299.1 1.39 41.2 November 533.2 1.20* 64.0 December 933. 0.96* 89.0 January 748.6 0.75* 56.1 February 404. 0.525 21.2 Total 3310.0 gal. 560.5 lbs. * estimated Table 5. "runoff" summary by months, 1958-59. Dry matter from "runoff" and "floor" effluent was ana-lysed for protein and ash (Table 6). Sample Date Oct. 13 Oct. 13 Feb. 4 Mar. 6 Description Runoff Floor Effluent Runoff Floor Effluent Dry Matter per 100 gal. 13.9 80.6 5.25 22.0 M x 6.25$ 31.4 27.2 26.0 30.8 Ash Percent 22.2 24.0 38.2 32.3 Balance 46.4 48.8 35.8 36.9 Table 6. Protein and ash i n dry matter from "floor" effluent and "runoff", 1958-59. The total observed dry matter losses, August 1958 to February 1959, in the effluent liquids are shown i n Fig. 19 in the discussion. Silage samples were taken for observation during the -17-storage period both years. A summary showing sampling dates and physical characteristics of the samples i s given in Table 7. Sample Depth Dry Matter Density Dry Matter i n Inches Percent of (wet l b . per Wet Weight weight) cu. f t . l b . per cu. f t .  1956-57 Hov. 27 19.3 47.3* 9.2* Jan. 14 l i - 1 1 frozen 20.1 48,6 9.8 1957-58 July 29-30 pea vine 17.2 Oct. 14 19.0 Jan. 28 0-6 15.0* 20.0* 3.0* 6-18 14.3 27.8 3.97 18-30 15.2 39.7 6.03 30-42 15.0 57.5 8.61 42-52 16.7 45.0 7.50 1958-59 Average 15.3 39.7 6.05 July 29-30 pea vine 21.7 45.0** 9.8** Sept. 3 0-12 23. 11.0 2.38 12-24 21.4 34.3 7.32 24-36 19.3 48.1 9.1 36-41 21.6 43.2 9.1 Average 20.6 33.7 6.95 Jan. 31 0-12 17.7 13.8 2.44 12-24 16.9 34.0 5.75 24-36 15.5 44.6 6.91 36-40 17.2 55.1 9.45 av.3*** 16.4 37.9 5.40 av.2 15.7 37.3 5.83 Mar. 7 0-11 9.5 29.6 2.81 11-23.5 16.9 41.7 7.06 23.5 -35 15.5 44.4 6.69 35-39 15.8 49.2 7.79 Average 14.7 39.7 5.86 •estimated ••calculated from packed volume and total weight of packed silage, •••including samples from very soft, deeply rotted area Table 7. Pea vine and silage sample summary. -18-Proximate analysis of selected samples from 1956-57 and 1958-59 was done and the proximate composition observed i s li s t e d in Tables 8 and 9. Sample Date Nov.27,1956 Jan. 14» 1957 Percent Dry Matter 18.7 20.2 Density 471bs. per cu, . f t . 48.71bs, . per cu.ft. Dry Matter Analysis Protein 13.1 13.2 Pat 3.7 3.8 Fiber 30.7 30.4 Ash 11.8 10.3 N.F.E. 38.7 42.3 Silage Analysis (wet basis) Protein 2.3 2.7 Fat 0.7 0.8 Fiber 5.7 6.2 Ash 2.6 2.3 N.F.E. 6.9 8.7 Table 8. Proximate < composition of 1956-57 silage samples. Sample Date Sent.3.1958 Jan.31.1959 Mar.6.1959 Percent Dry Matter 20.6 16.4 14.7 Density 33.7 37.3 39.7 Dry Matter Analysis Protein 9.8 9.3 9.4 Fat 2.9 3.4 3.4 Fiber 24.6 25.8 26.4 Ash 17.5 17.5 18.0 N.P.E. 45.2 43.7 42.8 Silage Analysis (wet basis) Protein 2.01 1.52 1.38 Fat 0.60 0.56 0.50 Fiber 5.06 4.23 3.89 Ash 3.62 2.88 2.65 N.F.E. 9.34 7.17 6.30 Table 9. Proximate composition of 1958-59 silage samples from percolation test area. The silage depth decreased between packing and the end of the storage both seasons. In 1957-58, the silage settled about three inches in the f i r s t month, and a further three inches following a period of heavy rain Movember 10-14th. The total settling in 1957-58 was five to six inches. In 1958-59, con--19-siderably more settling occurred. Settling between August 1 and September 3 was approximately five inches, and from Sept-ember 3 to February 1 five inches, for a total of ten inches from f i l l i n g to feeding. The silage along the walls settled two to five inches more than the silage in the middle of the s i l o . Observations of the ensiled material after f i l l i n g sugg-ested that considerable losses occurred, both i n 1957 and 1958 (Fig. 13 and 14). Fig. 13. Surface conditions November, 1957. Fig. 14. Surface condition February, 1959. Metal piezo-meter tags can be seen on the silage surface. -20-Surface changes after f i l l i n g were slightly different i n 1957 and 1958. In 1957, the very wet pea vine packed tightly right to the surface, and a definite seal layer formed within a week. The exposed silage above the seal layer dried out somewhat, and the seal layer increased in depth for about three weeks. By September 1, the total "spoil" layer averag-ed about six inches deep. The "spoil" layer appeared almost impervious to water. Sur-face water from r a i n f a l l remained in depressions on the surface almost continuously from November 12th u n t i l the end of the test. Some surface mold and plant growth in the "spoil" layer started i n late November(Fig. 15)• Fig. 15. Surface condition, December, 1957. The f i n a l depth of the "spoil" layer varied from four to eleven inches and averaged about seven inches on January 28, 1958. The drier, less tightly packed pea vine ensiled in 1958 developed a relatively deep, incompletely rotted layer within a week after packing. This seal layer was about five inches thick, under a two inch surface layer of dried pea vine. This seal layer did not prevent some further decomposition of the silage under i t , and by September 3, the "spoil" layer was eight inches deep. A thin layer of very hot, rotting s i -lage was observed at the bottom of the "spoil" layer, where the temperature was 110°F. This rotting layer had cooled off completely by October 6, and appeared as a greyish paste lay-er about one half to one inch deep. The "spoil" layer depth was s t i l l eight inches October 6 and January 31. The final "spoil" layer depth March 6 ranged from five to eleven inches and averaged eight inches. As i n 1957-58, the "spoil" layer was very wet, and often had water lying in the surface depressions for several days after heavy rains. Mo plant growth was observed on the s i -lage surface. In both years another layer of material which could not be used developed under the "spoil" layer during the storage period. This "sub surface" layer was negligible i n early September, and the normal transition from "slimy" seal mater-i a l , through partly rotted, dark material, to light olive green silage usually took place i n one inch or less of depth. With Increasing storage time, a definite layer of darkened, sour or moldy smelling silage was observed, under the "spoil" layer(Table 10). This "sub surface" layer had a definite top boundary of partly rotted silage i n the "spoil" layer, but the lower boundary was less clearly defined. The colour and odour of tne material gradually changed to that of palatable silage over a depth of two to four inches, so the actual depths giv-en were arbitrary. 1957-58 1958-59 Date Percolation "Toe" Percolation "Toe" Test Area Area* Test Area Area** Sept. 1-3 0 0 0 1" Oct. 13 1" 4" 1" 1-3" Jan.28-31 3" 12" 4-8" 8" Mar. 6 6-8" 10" * 30 feet from percolation test area ** 20 feet from percolation test area Table 10. Thickness of the "sub surface" layer during the storage period. The investigation of saturation i n the "sub surface" lay-er using electric piezometers was unsuccessful, and no con-clusive saturation condition was revealed (Table B). Observation of the exposed end of the silage during feed-ing i n 1957-58 showed some liquid was draining from the vert-i c a l face of the silage. The liquid appeared to be coming primarily from the darker "sub surface" layer, and no measur-able flow occurred from the light-coloured silage. The c o l -lection of horizontal flow from the vertical face of the "sub surface" layer i n March 1958 showed a measurable flow occurr-ed. This flow was estimated to be not less than twenty-two gallons through the "toe" end of the "sub surface" layer dur-ing the storage period. Pairs of core samples for permeability tests were taken from near the floor of the s i l o . Vertical cores could be taken at any depth, but undisturbed horizontal cores could only be taken from the relatively dense, uniformly packed s i -lage. The f i r s t pair of core samples were from a section where the silage was tangled, and the horizontal core obtain-ed gave exceptionally high permeability values (Table 11). Core #1 #2 #3 #4 #5 Direction vert. horiz. vert. horiz. vert. Description paired with #2 #4 paired with #3and #5 Diameter 3.188" 3.188" 3.188" 3.188" 3.188" Length 6.125" 3.5" 5.&9" 5.69" 4.81" Head 1" 1/2" 1/4" 1/4" 1/4" Volume per hr. 24cc. 6720cc. 408cc. 1065cc. 37.5cc. Calculated K .10 30.4 2.3 6.0 .21 Density 46.1 44.2 45.3 Ratio % : % 304 2.6 29.6 Average EU £ _ 21*1 Average Ky *- L , x Table 11. Permeability of core samples, March 6,1959. The horizontal and vertical permeabilities, Kg and re-spectively were calculated from the relationship Q= A K i , where i , the hydraulic gradient = h, the head differential divided by L, the length of the sample through which the wat-er moves(,Appendix F). Some other observations were made after the s i l o was open ed to the cattle. In 1957, 1958, and 1959, the cattle showed strong preference for silage from some areas of the feeding face. Some silage, was rejected completely while adjacent s i -24-lage was eaten up to the l i m i t of a v a i l a b i l i t y (Fig. 16). F i g . 16. Appearance of exposed face, February, 1958. In most cases the rejected silage was s l i g h t l y darker i n colour, and appeared very wet. There was no discernable d i f -ference i n smell between acceptable and unacceptable silage i n most cases, but some rejected silage did smell s l i g h t l y moldy. The unacceptable areas ranged r i g h t from f l o o r l e v e l to the base of the "sub surface" la y e r . The rejected silage appeared somewhat l i k e the material i n the "sub surface" layer i n some places. The p o s i t i o n of the rejected silage areas was apparently random at any one time, but as the exposed face advanced dur-ing feeding, these areas tended to indicate i r r e g u l a r volumes along the length of the s i l o . The arrangement of the silage remaining i n the rejected areas suggested that these areas occurred between the separate fork loads of pea vine f i l l e d into the s i l o . - 2 5 -The f i n a l t otal dry matter yield i n 1957-58 was calcu-lated from estimated dimensions and observed density f i g -ures, since exact weight i n the test s i l o was not known. The maximum dry matter yield of palatable silage was calcu-lated as 35%, and the minimum as 25%, depending on the exact weight and average percent dry matter of the input material for the whole s i l o . The calculated "spoil" plus "sub sur-face" layer loss was estimated at between 6% and 8.5% of the total input dry matter. The calculated storage losses and yields, for the 1958-59 season are given i n Table 12. Weight of silage put into the s i l o ........ 130,5501bs. Dry matter 21.7% of wet weight. Total dry matter input 28,3001bs. Sept. 3,1958 Mar. 6,1959 Pounds Percent Pounds Percent of input of input Total dry matter 11,800 41.7 9,700 34.3 loss of dry matter 2,100 7.4 Dry matter i n "spoil" layer 830 2,9 1,010 3.6 Dry matter i n "sub surface" layer 110 0,4 2,400 8.5 Net edible dry matter 10,860 38.4 6,290 22.2 Loss of edible dry matter dur-ing storage period 4,570 16.2 Table 12. Calculated storage losses and yields, 1958-59. Observation of grass silage in an adjacent s i l o with one half i t s length covered by a plastic sheet and held down with four inches of sawdust were also made. The surface loss was negligible under the plastic sheet but amounted to about six inches in the unprotected silage. However, the total depth of silage adjacent to the plastic sheet was two inches less than the depth under the sheet, and the total dry matter per square foot of surface was only five sixths as great as the total dry matter per square foot under the plastic sheet (Table 13). Silage from Silage from protected area exposed area Percent dry matter 21.2 19.4 Average density pounds per cu.ft. 37.6 35.0 Dry matter per cu.ft. in pounds 7.96 6.78 Dry matter, pounds per sq. f t . of surface 35.4 31.6 Surface loss, pounds per sq. f t . of surface 0 1.7 Net dry matter, per sq. f t . of surface 35.4 29.9 Comparative yield of dry matter 6 : 5 Table 13. Percent dry matter, surface loss, and yield for grass silage (a) exposed and (b) covered by a plastic sheet, Janu-ary, 1959. One effect of the plastic sheet was undesirable. Since most of the r a i n f a l l ran towards the wall of the s i l o and the plastic sheet was not carried up the wall at a l l , a large volume of water was spilled into the silage near the wall. A definite loss by molding and rotting occurred along the walls. This can easily he prevented by extending the plastic sheet -27-up the wall about six inches. The cost of plastic sheets for use as si l o covers was $,02 per square foot. The f i f t y foot by thirty foot sheet which was used i s i n f a i r l y good condition, and can be reused after repairs estimated at $5.00, Its useful l i f e i s estimated as two or three years. The cost of applying, covering, and removing the plastic sheet were not known, but were estimated as comparable to, or slightly less than, the cost of stripping the "spoil" and "sub surface" layers from pea vine silage. Pea vine cost $5.00 per ton delivered, to the s i l o . The total pea vine ensiled i n 1957-58 was 397 tons i n two si l o s , and i n 1958-59 65.2 tons.. DISCUSSIOM Samples taken by the core sampler were used whenever possible to provide data for density calculations. The den-sities so obtained are probably less accurate than cubic foot examinations, since the samples are smaller and more subject to local variation. Otis and Pomroy reported their four inch core sampler gave density values slightly lower than cubic foot sampler values (14). The fact that impact must nearly always be used suggests some error may be intro-duced into both density and moisture determinations by wat-er loss caused by the impact pressure. Por this reason a l l samples were taken with as nearly identical procedure as possible, and cores were removed after each foot of depth had been penetrated as estimated from sampler penetration and checked by measurement of the hole depth. Errors in -28-surfaee density due to trampling etc. were not compensated for, but a minimum of inspection and sampling i n the perco-lation test area was done from September to February to keep these errors to a minimum. The density calculated from cores from the deep sampler (Fig. 3) was compared to the density of short cores, from the same location made with the permeability corers and show-ed good agreement, 44.4 and 45.2 lbs. per cubic foot respect-ively. The "floor effluent" collected from the "floor" area each season was very small compared to total r a i n f a l l (Fig. 17 and 18). In 1957 the very wet material lost two thirds of the total "floor effluent" collected in the f i r s t month, before any significant r a i n f a l l occurred. The relatively dry pea vine ensiled i n 1958 did not show an appreciable liquid loss u n t i l after at least two inches of rain had fallen on the s i -lage surface.. The relatively small loss of liquid from the floor drain-age system i s emphasized when the total collected "floor ef-fluent" for 1957-58 over six months, was equivalent to 0.17 inches or .55$ of the total r a i n f a l l . This compares to 0.31 inches or 0.8 per cent of the total r a i n f a l l i n the 1958—59 season when a slightly shallower silage mass was present. In 1957-58, the amount of "wall effluent" collected was negligible after August 3 . The amount collected in 1958-59 was appreciable, and showed a flow rate similar to the rain-f a l l graph u n t i l October 17th. Then and periodically during - 2 9 -the remainder of the storage period the "wall effluent" flow became erratic and unpredictable. Apparently an a i r lock formed somewhere in the wall drain line, perhaps due to gases produced by fermentation i n the liquid during periods of slow drainage. Efforts to blow out the line by a i r pressure and to f i l l up the lines with water did not prevent recurrence of the apparent blockage. After the line was dug up and re-buried i n December the "wall effluent" flow was more even, but never attained the high values observed i n early October. It i s possible that overflow from the wall area may have influenced the "floor" effluent volume, but since relative-l y small volumes are involved, no serious errors should re-sult even i f some interchange occurred. There was some l i -quid loss through the wooden walls of the s i l o . The silage surface was sloped more steeply towards the sides than to-ward the low end of the s i l o , and the silage depth along the walls was uneven. Considerable surface runoff went to the wall and was unable to run parallel to the wall towards the "toe" and "runoff" collection lines. Some leakage was ob-served along the unbacked north wall, but no leakage was evident through the earth hanked south wall. Ho estimates of this loss were made. The "runoff" and end guard area drains were combined since guard area volume was assumed to be too small to af-fect the "runoff" volume appreciably. The observed "runoff" was related to r a i n f a l l , but only accounted for from 30-49 percent of the observed r a i n f a l l (Table 14). I i i > o 17, R a i n f a l l and F l o o r E f f l u e n t C o l l e c t e d , by Months, 1957-1958 0 1 1 1 1 1 1 Augo Septp Oct. Nov. Dec. Jan. Feb. F i g . 1 8 . R a i n f a l l and E f f l u e n t C o l l e c t e d , b y Months, 1 9 5 8 - 1 9 5 9 . Rain i n E f f l u e n t Inches i n G a l l o n s Augc Sept. Oct. Nov. Dec. Jan. Feb. -32-Month Rainfall Total effluent Percent of in gallons in gallons r a i n f a l l collected August 490 156 31.8 September 550 270 49.1 October 1130 425 37.6 November 1940 592 30.5 December 2160 946 43.8 January 1805 782 43.4 February 1530 479 31.1 Total 9605 3650 38. Table 14. Rainfall and total effluent collected by months, 1958-59. Ponding i n the dam at the "toe" of the s i l o tended to aver-age out the daily r a i n f a l l readings by allowing a carryover from day to day or while the tanks were being drained. Drain-ing had to be done by hand pump or compressed a i r , when the water level i n the f i e l d drains was high enough to prevent gravity drainage of the tanks. Maximum capacity of the run-off collection system was about 130 gallons per day for three f i l l i n g s and three drainings at six hours per cycle.. This was equivalent to about 1.0 inches of rain in two days. At the start of the winter this was f e l t to be adequate, as the surface roughness of the s i l o delayed surface runoff consid-erably. However, this capacity was found to be inadequate for the highest observed ra i n f a l l s , and estimates of the vol-ume flowing over the dam were made to compensate for this loss. No calculations of evaporation loss were made because low mean temperature and high humidity prevailed through most of the storage period and evaporation losses were there-fore assumed to be small. A liquid balance for the 181 day storage period from Sept-ember 3, 1958 to March 6, 1959 revealed a 4541 gallon short-age in the total volume of liquid measured compared to calcu-lated volume of water i n r a i n f a l l on the s i l o for the same period. This unmeasured liquid was 49.8 percent of the total r a i n f a l l during the storage period (Table 15). Total r a i n f a l l 34.34 inches 9115 gal. Total in effluent 3494 gal. Retained in silage 1080 Total measured 4574 Not measured 4541 Total 9115 gal.9115 gal. Percent of total r a i n f a l l not located 49.8%, Table 15. Calculated liquid balance September 1958-March 1959. A part of the liquid loss was overflow from the collect-ion tanks. Other small losses, occurred by wall leakage, but these were not considered to be appreciable. The main source of error i n the liquid collection system was the inadequate capacity of the "runoff" measuring system. Sustained heavy rain totalling more than 1,0 inches per forty-eight hours overloaded the "runoff" collection system so that a large part of the ensuing runoff overflowed and was lost . Por this reason, the total uncollected liquid was assumed to be "run-off" . The dry matter losses in the effluent from September 3, 1958 to March 6, 1959 are summarized in Table 16. F i g . 19. Dry Matter Losses i n Effluent by Months, 19b7-58 and 1956-59 200 Monthly Dry 150 Matter Losses In LOO Pounds 50+ Dry Matter in"Wall" E f f l u e n t Dry Matter in"Runoff" Dry Matter in"Floor" Effluent v • /. //A /57 /58 /57 /58 /57 /58 /59 /58 /57 /58 /57 /58 Aug. Sept. Oct. Nov. Dec. Jan. Feb, /58 l I -35-Effluent Dry Matter Percent of in Pounds Input Dry Matter Floor effluent 31.4 0.11 Wall effluent 183.3 0.65 Runoff (measured) 406.5 1.45 Total calculated dry matter 621.2 2.21 Runoff (estimated) 650. 2.3 Total estimated dry matter loss in ef-fluent 1271.2 4.5 Table 16. Summary of dry matter losses i n effluent liquids, 1958-59. The total dry matter in the measured effluent during the storage period was 621.2 pounds, or 2.2% of the input dry matter. The estimated liquid not collected, at the October dry matter content of 1.39%, was calculated to contain 650 pounds dry matter or 2.3% or input dry matter. The correct-ed dry matter loss for the storage period was 1271 pounds or 4.5%. The dry matter losses i n the "floor" effluent were for an area of 210 square feet of the 504 square feet of the tot-a l floor area. The"wall"effluent came from 30 feet of the 72 foot wall length. The corrected losses for the whole "floor" and "wall" areas respectively, were 0.27% and 1.55% of the input dry matter. These corrections do not affect the total dry matter loss as this dry matter was theoretic-ally collected in the "runoff" l i q u i d . Total estimated dry matter losses i n effluent from August to February inclusive were 1286 pounds or 4.55 percent of input dry matter. -36-The protein and ash in the "floor" effluent and "runoff" was considerably higher than the reported protein and ash for effluent from tower silos (9). The protein remained f a i r l y constant at 26-31% during the storage period, and was lowest in the dry matter from "runoff" liquid in February (Table 6). This i s above the 17-28% range given for dry matter in e f f l u -ent from tower s i l o s . The percent ash i n dry matter in Oct-ober was nearly equal at 22.2 and 29.0% in "runoff" and "floor" effluent respectively, and near the top of the range of 13-23% reported for dry matter in effluent from tower s i l o s . At the end of the storage period the ash content was 38.2 and 32.3% respectively for the dry matter in "runoff" and "floor" e f f l u -ent. The total loss of protein in the effluent dry matter was estimated as 372 pounds or nearly 9% of the estimated 4200 pounds protein in the pea vine. For ash, the total loss in the effluent was estimated as 370 pounds or 18% of the estimated 2100 pounds ash in the pea vine. This relatively important loss was more important than the small total dry matter loss of 4.5%. This was similar to reports by Watson (19) that 4% of mineral and 3.5% of protein were contained in less than 1.5% of the total dry matter lost as effluent from a tower s i l o . It i s interesting to note that the "runoff" from the end of the sil o September 3rd had very nearly the same dry matter eon-tent, 10.0%, as the "wall" effluent, 10.7%. The "floor effluent" was insufficient for sampling. This suggests that l i t t l e dry matter loss due to rain had occurred prior to September 2rd. This contrasts with the percent dry matter on October 13, when "run--37-off" contained 1.39 percent, "wall" effluent 6.06 percent and "floor" effluent 8.06 percent dry matter after approxi-mately four inches of rain had fallen onto the silage. The distribution of the dry matter losses i s indicated in Fig. 19. The very high dry matter losses i n September "runoff" are due to the high observed percent dry matter, and moderate volume, not to a large volume of effluent. The November, December, and January "runoff" dry matter came from larger volumes of effluent and much lower percent dry matter contents. The silage samples taken at different times over the stor-age period showed some changes in composition with time. Dur-ing 1958-59 the relatively small decreases i n protein, and N.F.E., and relatively small increases i n fat, fiber, and ash on a dry basis were not as important as the general decrease on a wet basis of a l l constituents with increasing water con-tent over the storage period. The reduction i n the quality of the silage was primarily a reduction in the dry matter available per one hundred pounds of silage, which decreased from 20.6 pounds on September 3rd to 14.7 pounds on March 6th. The rise i n percent dry matter observed in 1956-57 was due at least i n part to an ice layer on the surface of the si l o for some weeks, so that l i t t l e water was able to get into the silage, and probably i n part to variation i n the material. The silage samples also showed some changes in density with time and depth. The observed density of the silage i n -creased with both increasing depth and increasing time (Ta-ble 7). In a l l cases the surface density was the lowest. -38-The increase in density with depth was associated with high-er dry matter content per cubic foot in almost a l l cases. However the increase in density with time is primarily an increase i n water content, with the actual average dry matt-er i n pounds per cubic foot decreasing with time. The aver-age dry matter per cubic foot could be expected to decrease slightly, as the decrease in volume over the storage period was relatively small, and the calculated dry matter losses in the total collected effluent liquids totaled 0.35 pounds per cubic foot of silage. The storage losses observed were over a storage period of about 180 days which i s comparable to farm practice. The most prominent loss was in the "spoil" layer, but in fact this loss was almost constant over the storage period (Table 12). The increase i n the surface loss over the storage per-iod as a percent of the total input dry matter was only 0.7 percent i n 1958-59. The total "spoil" layer loss on March 6th was 3.6 percent as compared to 2.9 percent on September 3rd, only one month from f i l l i n g . This layer showed consid-erable variation i n depth from area to area, but the varia-tion did not form any predictable pattern. The most important dry matter loss during the storage period was i n the "sub surface" layer. This layer was neg-li g i b l e or indistinguishable on September 3rd, but amounted to 2400 pounds or 8.5 percent of the dry matter input by March 6, 1959. The storage losses from September 3rd to March were "sub surface" layer 8.5 percent, observed dry matter in effluent 2.2 percent, estimated dry matter i n -39-unmeasured effluent 2.3 percent, and "spoil" layer loss of 0.7 percent. The total exposure loss directly attributable to exposure for 180 days was 8.1 percent in "sub surface" layer, 7.4 percent in dry matter lost from the silage, and 0.7 percent in surface loss or 16.2 percent of input dry matter. The 1957-58 observations suggested that the "sub sur-face" layer was formed by liquid from the surface contamin-ating the silage under the "spoil" layer with decomposition products, and perhaps changing the conditions of aeration, water content, and even acidty to the point where secondary fermentation occurred. The depth of the "sub surface" layer increased toward, the "toe" of the s i l o , and was least in the area around the highest point of the si l o surface. The thickness of this layer also increased with time, but the thickness at the "toe" end of the s i l o was always greater than at the upper end. The 1958-59 observations are similar to the 1957-58 ob-servations. The "sub surface" layer developed only after at least four inches of rain had fallen, and increased i n depth with time, and toward the "toe" of the s i l o . It also i n -creased in thickness slightly toward the walls, as the slope toward the walls was greater than the slope toward the "toe" of the s i l o . Relatively large variations i n thickness occ-urred. One very dense hump in the s i l o had less than three inches of "sub surface" layer loss, while a low area along the wall near the "toe" had nearly eighteen inches loss for several feet. The development of the "sub surface" layer -40-appears to be related to the amount of water passing through the "spoil" layer and into the silage, and to the mode of wa-ter movement. The use of electric piezometer cells to try to detect saturation of the mass, particularly the "sub surface" layer, appears to he a useful tool for further moisture studies, a l -though they were not too satisfactory i n this study. The principle i s satisfactory and responses are fast (Table E, appendix). Failures were apparently caused by minute cracks in the plastic insulation on the wires, which allowed the silage liquid to contact the wires. The small area of contact and of conducting f l u i d kept the reading much lower than that of a submerged c e l l , but allowed a small "leak" reading. The ammeter readings were related to the distance between the test c e l l and the ground rod, but this variation was small and did not interfere with other readings. The area of electrode in contact with the silage i s more important than the spacing from c e l l to ground electrode, but this area could be easily increased i f necessary. Scale readings decreased f a i r l y rapidly after about two seconds of current flow, apparently as a result of polarization of the elect-rodes. The original maximum reading could be duplicated after waiting about two minutes, or by reversing the polar-ity of the system, or by very short successive reading i n -tervals. This " f a l l i n g off" of the scale readings did not interfere with separation of c e l l saturation and "leak" readings. -41-Although the piezometer cells did not indicate saturation at the twelve inch level, a measurable volume of effluent was collected from the vertical face of the "sub surface" layer at the "toe" end of the s i l o . This estimated flow, 22 g a l l -ons over 180 days, is very small, and is hardly enough to account for the spoiling of 2400 pounds of silage. It i s probable that much higher flow rates exist at certain places over the vertical face, and that some higher flow rates may be observed at different times as r a i n f a l l varies. This small flow, was calculated from the very conservative assumption that the flow rate observed was twice the average flow rate, and that the average conducting area of the "sub surface" layer was one half the f i n a l end area of fifteen square feet. This flow was nearly as great as the total flow from the two hundred and ten square foot "floor" area. Since the possibility that appreciable horizontal flow occurred had been suggested in 1957 by observations at the exposed vertical face, some tests of horizontal and vertical permeability were done. The values of permeability i n a horizontal direction (Kg) were in every case higher than the values for vertical permeability (Ky.) through the mass (Table 11). The f i r s t pair of samples gave extremely different values of K^ and K^ .. The very high K^ was probably due to channelling i n the horizontal core sample, which was d i f f i c u l t to get with-out damaging the silage structure. The next set of samples s t i l l had a considerable variation between the Kg values. A l l the Kjg; values were larger than the largest K^- value. The -42-average K^ to average Ky ratio i s 21 : 1. Since the K value i s a measure of the rate of flow of water through a mass, a high K value indicates high possible flow rates. The Kg to Kg ratio indicates that water w i l l flow horizontally more readily than i t w i l l vertically under the same hydraiulic head. The development of the "sub surface" layer, which increases i n depth with increasing storage time, and also increases in depth towards the "toe" or low end of the s i l o can apparently be explained i n part ( by the Kg ^ Kg ratio. Considerably more work could be required to es-tablish a definite relationship between the shape of the "sub surface" layer and the observed Kg to Kg ratios. The cores used for the K determinations were taken from the thirty inch depth, since this was the only location at which a satisfactory horizontal core could be obtained. The % to Kg relationship i s probably less definite as the den-sity of the silage decreases, and may not, in fact, be de-terminable by this technique in the "spoil" and "sub surface" layers. A dry matter balance from September 3 to March 5 showed f a i r agreement (Table 17) Calculated Dry Matter in PoundP  September 3, 1958 11,800 March 6, 1959 9,700 Total estimated dry matter in effluent 1.271 11,800 10,971 Not accounted for 829 Table 17. Calculated dry matter balance, September 1958 to March 1959. -43-The calculated dry matter loss from the si l o over the stor-age period, as given by the difference i n dry matter i n the September 3 and March 6 samples, was 2100 pounds or 7.4 per-cent of the input dry matter. The dry matter not accounted for in the effluent during the storage period, was 829 pounds, or 2.9 percent of the input dry matter. There i s no way of demonstrating whether the whole 2100 pounds of calculated dry matter lost from the silage was a l l in the effluent l i -quids or whether some dry matter was lost by surface oxida-tion and decomposition. The presence of the very hot, almost ash like layer under the "spoil" layer i n September suggests surface losses may have contributed to the storage dry matt-er loss. 1957-58 observations showed a moderate "spoil" plus "sub surface" loss during storage and a negligible dry matter loss in liquid collected from the "floor" area. More data on dry matter content in the si l o i n September 1957 would be needed to accurately measure the dry matter losses. The maximum net yield and the maximum dry matter at the end of the stor-age period, as estimated from photographs and other data, was 35 percent and 40 percent respectively, of the total dry matter input (Fig. 20). The net dry matter yield from the 65 tons of pea vine s i -lage was approximately 6290 pounds or 22,2 percent of the total dry matter delivered to the farm. The net dry matter yield on September 3 was 38.4 percent of the gross dry matt-er input. Total losses i n the f i r s t month by fermentation and surface losses were calculated as 61.6 percent as com-100 T Percent of 80 Input Dry Matter 60 40 --£0 --/57 /5c At F i l l i n g F i g . HQ. Dry Matter D i s t r i b u t i o n and Y i e l d , 1957-1958 and 1958-1959. Estimated MaXo T o t a l Dry Matter /57 / 5 8 Sept.1-3 ' \ \ \ " S p o i l " Layer "Sub s u r f a c e " Layer Net E d i b l e Dry Matter l I /58 /D 9 J an. Feb, pared to a storage loss of 16.2 percent for the remaining 180 days. The largest part of the total loss from f i l l i n g to feed-ing took place i n the f i r s t month after f i l l i n g . The subse-quent storage losses were much less important, although they amounted to over 40 percent of the yield. In 1958-59 the total cost of a l l loss at $5.00 per ton pea vine delivered, no labour included, was $253» compared to the original cost of $325. The cost of the total loss per square foot of silage surface was 46.8 cents. The i n i -t i a l cost of the plastic sheet was 2 cents per square foot. The loss i n the "sub surface" layer alone was approximately 6.1 cents per square foot of exposed surface for the stor-age period. Storage losses were estimated at 12.3 cents per square foot of exposed surface from September 3 to March 6th. The effect of a plastic cover over part of the mature grass silage in the adjoining s i l o indicated much lower losses from the plastic protected area. The calculated "spoil" layer loss from the exposed sample was 6.7 percent of the total dry matter present. The apparent effect of the plastic was to reduce the surface loss almost to zero, and also to reduce the dry matter lost by fermentation, as re-ported for experimental metal silos (7). The combination of higher density and higher dry matter content per pound of wet silage under the plastic sheet as compared to the adjacent area resulted in a net .yield from an adjacent unprotected area. The exposed silage contained approxi-mately 85 percent of the net yield of the protected area i n almost the same volume, made from the same material at the same time, under similar conditions' of packing and sprink-l i n g . The plastic sheet was applied shortly after packing and covered with sawdust to hold i t down and protect i t . The depth of silage in a horizontal s i l o affects i t s storage efficiency considerably. In general, the e f f i c i -ency of storage increases as the depth increases, since both unit building cost and percent loss of silage decrease with increasing depth. The net yields at shallow depths are very low, and for most outdoor storage any f i n a l settled depth under four feet is li k e l y to be disappointing. The B.C. De-partment of Agriculture Field Crops Branch i n a bulletin issued March 1958 recommend a minimum width of 16 feet and a minimum settled depth of 6 feet for horizontal s i l o s . Surface protection against r a i n f a l l by any means during the storage period would have to be done for a small cost, much less than would provide a roof, and amounting to less than $34.00 for a 36 foot by 15 foot silage surface area. A plastic sheet would more than pay for i t s e l f under such conditions, but no roof could be constructed and maintained for this cost. The roof would prevent most of the "sub sur-face" loss and the runoff loss, but would not affect the large losses in the fermentation period, nor the develop-ment of the "spoil" layer. A plastic sheet to seal the surface of the silage can be justified for lower Fraser Valley f a l l and winter exposure -47-conditions, on the basis of reduced storage losses alone. A considerably greater feeding yield could be expected i f the plastic sheet was used to seal the silage surface immediate-ly after packing was completed. In practice, plastic sheets used to seal horizontal silos and stack silos immediately after packing have resulted in total dry matter losses which are negligible under English exposure conditions (4, 16). A further unmeasured value of both a roof and an imper-vious surface cover is the maintenance of the original per-cent dry matter with the associated higher dry matter i n -take by animals self feeding from the silage. Hash consid-ered that any silage with dry matter under twenty percent was low dry matter silage (18). CONCLUSIONS The total loss of dry matter i n pea vine silage in an un-protected horizontal s i l o from f i l l i n g to feeding was very high. The losses during the f i r s t month after f i l l i n g were the most important, and any method of reducing the losses dur-ing the fermentation period would be more important than pro-tection during subsequent storage. Storage losses increased with time and r a i n f a l l . The major storage loss resulted from the formation of an unpalatable "sub surface layer" under the "spoil layer", apparently from translocation of material from the "spoil layer" and some secondary fermentations. The dev-elopment of the "sub surface" layer was apparently related to liquid movement through the "spoil layer" into the silage. A second important loss was the disappearance of dry matter -48-from the silo during the storage period. This latter loss was greater than the calculated dry matter loss in a l l the e f f l u -ent liquids for the same period. The dry matter lost in a l l effluent liquids during the st-orage period was only 4»5 percent of the input dry matter, but i t contained an estimated 9 percent of the protein and 18 per-cent of the ash. The greatest part of the dry matter loss in the effluent liquids was in the"runoff"liquids collected from the "toe" of the s i l o . The dry matter lost in liquid pass-ing down along the walls was small and the dry matter liquid draining from the floor of the silo was negligible. The density of the silage increased with time but the dry matter per cubic foot and dry matter per pound of silage de-creased during the storage period. The use of a plastic sheet to seal the top of the silage immediately after packing appears to be the best and most ec-onomical method of reducing the dry matter losses during storage. The cost of the plastic and handling costs are much less than the value of the silage conserved by the sheet dur-ing the storage period. A further advantage when the sheet is used to seal the silo immediately after packing, are the lower fermentation loss and the maintenance of the dry matt-er content per pound of silage. -49-LITERATURE CITED 1. Allred, K.R. & Kennedy, W.K., "The Use of Small Silos to Determine Dry Matter Loss During Ensiling." Agronomy Journal, Yol. 48, Pages 308-313, July, 1956. 2. Archibald, J.G. & Gunness, C.I., "Seepage Losses Prom a Silo." Journal of Dairy Science, Yol. 28, Pages 321-324, April 1945. 3. Barnett, A.J.G., "Silage Fermentation." Butterworths Scientific Publications, London 1954. Drainage, Pages 57-62. 4. Beaucourt, D.F., "Plastic Silos, A Mew Way to Make Silage From Grass Crops." Implement and Tractor Trade Journal, 1957 (Abstract). 5. Blish, M.J., "Factors Affecting Quality and Composition of Sunflower Silage." Montana Agricultural Experimental Station Bulletin Ho. 74, 1921. 6. Bobsledt, G., "Hutrient Yalues of Hay and Silage As Affected By Harvesting, Processing and Storage." Agricultural Engineering (Journal of American Society of Agricultural Engineers) Yol. 25, Pages 337-340, Sept. - Oct. 1944. 7. Cooper, D.J., Gordukes, W.E. & Kalbfleisch, W., "A Test Silo and Apparatus For Ensiling Studies." Canadian Journal of Plant Science Vol. 38, Pages 181-183, April 1958. 8. Esmay, M.Li , Brooker, D.B. & McKibben, J.S., "Lateral Pressures on Horizontal Silos." Agricultural Engineering (Journal of American Society of Agricult-ural Engineers) October 1955, Vol. 36, Pages 651-653. 9. Le Clerc, J.A., "Losses in Making Hay and Silage." U.S. Agricultural Yearbook 1939. Pages 992-1016. 10* "Making Good Silage in Horizontal Silos." Agricultural Research Report Vol. 6 No. 8, Feb. 1958, Page IE U.S. Department of Agriculture, Washington, D.C. 11. Monroe, C.F. & Others. "Losses of Nutrient in Hay and Meadow Crop Silage During Storage." Journal of Dairy Science, Vol. 29, Pages 239-256. Bibliography Pages 252-256, April 1946. 12. Moore, H.I., "Silos and Silage." Farmer and Stockbreeder Publications, London 1950. 13. Otis, C.K. & Lieu, R.C., "Here's Where You Are Getting Silage Losses." Minnesota Agricultural Experimental Farm and Home Service. Vol. 13, Pages 12-13, 1955. -51-14. Otis, O.K. & Pomroy, J.H., "Density, a Tool in Silo Research." Agricultural Engineering (Journal of American Society of Agricultural Engineers) Mov-Dec. 1957, Yol. 38, Mo.. 11-12, Page 806. 15. Perkins, A.E., "Losses i n Silage Making." Ohio Agricult-ural Experimental Bimonthly Bulletin. Yol. 220, Pages 32-34. Jan. 1943. 16. "Silage With a Plastic Skin", Power Farming i n Australia and Mew Zealand, 1957 (Abstract). 17. S i l l , J.T. & Sears, R., "Roofing a Silo to Reduce Stor-age Losses." Mew Zealand Journal of Science and Technology. April 1942. 18. Waterson, H.A. & Mash, M.J., "Silage Making i n Scotland" Scottish Agriculture, Yol. 36, Mo. 1, Summer 1956. 19. Watson, S.J., "The Science and Practice of Conservation Grass and Forage Crops." The F e r t i l i z e r and Feeding Stuffs Journal, London 1939. Yol. 1 and 2. 20. Watson, S.J. & Ferguson, W.S.., "Losses i n Dry Matter and Digestive Mutrients i n Low Temperature Silage With and Without Molasses and Mineral Acids." Journal Agricultural Science, Yol. 27, Page 270107, Jan. 1937. APPENDIX TABLE A RAIMFALL 1957-1958 Month & Precipi- Month & Precipi Date tat ion Date tation Aug./57 Dec./57 22 .20 1 .17 23 .09 2 .16 Month total .29 3 .02 Cumulative .29 4 5 .25 .52 Sept./57 6 .05 1 .04 7 .04 5 .80 10 .05 6 .17 11 .09 17 .04 13 .37 26 .11 15 .33 Month total 1.16 16 .20 Cumulative 1.45 17 18 .28 .38 Gct./57 19 .08 7 .01 20 .26 12 .73 21 .09 13 .04 22 .11 22 .40 23 .48 23 1.02 24 .31 24 .20 25 .50 25 .20 26 .60 28 .20 27 1.12 29 .44 28 .11 Month total 3.24 29 .15 Cumulative 4.69 Month total Cumulative Hov./57 9 .05 10 .44 11 .91 12 .33 13 .59 14 .11 15 .12 17 .24 18 .01 21 .05 22 .04 23 .05 25 .12 27 .33 29 .09 30 .29 Month total 3.67 Cumulative 8.36 Month & Date -54-TABLE A (cont'd) RAINFALL 1957-1958 Precipi-tation Jan./58 1 .05 2 .11 3 .17 7 1.01 8 .22 9 .63 10 .33 11 .47 13 .90 14 .86 15 .36 16 .63 20 .17 21 .27 22 2.07 23 1.87 24 .18 25 .06 26 .04 27 .27 28 .76 29 .22 30 .02 Month total Cumulative 11.25 26.33 Month & Date Feb./58 3 5 6 7 8 9 12 14 15 16 17 18 20 21 22 23 24 25 Month total Cumulative Mar./58 4 6 7 8 Month total (to 15th) Cumulative Precipi-tation. .14 .02 .25 .35 .40 .41 .23 .12 .54 .27 .57 .15 .08 .59 .17 .43 .45 .01 .04 .32 .06 .12 5.77 32.10 .54 32.64 TABLE B RAINFALL 1958-1959 Month & Precipi- Month & Precipi-Date tation Date tation Aug./58 Nov./58 .10 2 .37 1 6 .23 2 .06 26 .02 3 .61 28 .94 5 .51 31 .30 7 .40 Month total 1.86 9 .03 Cumulative 1.86 10 .53 Sept./58 11 .50 8 .04 13 .01 10 .02 17 .52 15 .59 18 .68 16 .36 19 .56 17 .03 20 .09 18 .44 21 .53 19 .07 22 .06 20 .29 23 .14 24 .27 30 1.96 25 .04 Month total 7.29 26 .01 Cumulative 15.57 Month total 2.16 Dec/58 Cumulative 4.02 1 .53 0ct./58 2 .38 6 .09 5 .26 7 .15 6 .96 8 .13 7 .13 9 1.15 8 .24 10 .28 9 .38 11 1.18 10 .31 12 .18 11 .17 14 .07 14 .04 17 .46 15 .09 18 .17 16 .36 30 .06 17 1.45 31 .34 19 .28 Month total 4.26 20 .64 Cumulative 8.28 21 .03 23 .01 24 .38 25 .26 26 .21 28 .43 29 .48 30 .04 31 .05 Month total 8.11 Cumulative 23.68 Month & Date -56-TABLE B (cont'd) BAINPALL 1958-1959 Precipi-tation Month & Date Precipi-t a t i o n ^ Jan./59 4 .25 5 .42 6 .25 7 .16 8 .08 9 .29 10 .04 11 .33 12 .05 14 .09 15 .90 16 .14 17 .13 18 .02 21 .68 22 1.63 23 .58 24 .26 25 .03 26 .44 28 .01 Month total Cumulative 6.78 30.46 Feh./59 1 .01 2 .02 3 .04 4 .20 5 .44 6 .20 12 .11 13 .33 16 .36 19 .18 22 .18 23 .43 24 2.01 25 .08 26 .05 27 .93 28 .07 Month total Cumulative 5.74 36.20 Date -57-IABLE C EFFLUENT PLOW PROM "FLOOR" AREA P e r c e n t 1 9 5 7 - 1 9 5 8 Gallons Dry Matter  Gallons Per Month July 31 Aug. 2 4 7 10 Sept.l 14 25 28 Oct. 14 26 Nov. 4 13 15 20 Dec. 2 9 21 Jan. 1 6 14 21 Feb. 3 3.72 4.05 5.27 2.00 1.35 1.78 1.98 1.98 5.60 JZT .25 .25 .25 •50 .10 1.00 .50 .75 1.35 .25 .25 14.69 0.75 0.75 1.60 2.30 .50 Total 20.59 -58-TABLE D RUNOFF AND EFFLUENT LOSSES 1958 - 1959 Date Effluent Effluent Effluent from from from Test Area "Wall" Area Guard & "Runoff 1 .05 gal. .75 gal. 10.0* gal. 3 .65 3.7 5 12.0* 6 .05 14.6* 7 2.5 (plugging) 8 22.5 9 10.0 10 23.0 11 20.0* 12 3.5 13 7.1 14 1.5 17 3.5 21 4.0 24 8.0* 26 4.5 29 8.2* L total .10 1.40 154.6 ,1 .12 10.0 2 .05 8 1.00 .70 39.8(airlocked) 15 1.00 0.7 16 .83 24. 17 .83 32. 18 1.00* 17. 19 2.00* 1. 19. 20 1.00* 29. 22 2.00* 34. 23 17. .5 24 .5 .5 2. 25 1.0 8.. 26 .5 7. 28 2. 3.5 30 l ? - 5 Aug. Month total 29.3 3.2 237.5 * some losses Volumes recorded are total accumulation since previous readings. -59-TABLE D (cont'd) RUNOFF AID EFFLUENT LOSSES Date 1958 - 1959 Effluent from Test Area  Effluent from "Wall" Area Effluent from Guarda*Runoff* 2.0 2.0 .5 43" Oct. 7 8 9 10 11 . 13 14 15 16 17 19 21 22 24 26 28 30 Month total Nov. 1 2 3 4 5 7 9 10 11 12 14 16 17 18 19 20 22 23 24 25 26 27 29 30 Month total * some losses Volumes recorded are total readings. .25 .125 .06 .06 .125 .06 .06 .125 .06 .06 7985 gal. .5 1.0* 45.0* 15. 27. 30. 2.5 (lin : 121.0 3.5 22.0* 30. 40. 40. leaned)40.* 45* 8. 2.6 1. 17. 12. 9. 6. -2. 5. 16. 299.10 2.5 13. 7.1 2.0 .75 15.5 9.2 2.6 5.2 57.85 gal, 6, 18. 10. 25 18. 21. 17. 15. 25. 5.2 5.0 25. 10. 55. 55.* 58. 28. 25. 40. 14. 12.5 9. 2.5 ?4 * 533.2 gal. accumulation since previous -60-TABLE D (cont'd) EUNOFF AND EFFLUENT LOSSES 1958 - 1959 Date Effluent from Test Area Effluent from "Wall" Area Effluent from Guard&Eunoff 1 80. 2 32. 3 .06 1. 32. 4 21. 5 .06 1. 26. 6 .06 30. 7 .12 2.5 50. 8 .06 32. 9 1. 18. 10 21. 11 18. 12 1. 12. 13 .06 14. 14 11. 15 6. 16 1. 6. 17 .06 64. 18 61. 19 35. 20 46. 21 1. 35. 22 .06 28. 23 31. 24 16. 25 31. 26 1. 26. 27 .06 21. 28 .12 1.5 28. 29 .06 41. 30 .06 1. 35. 31 26. L total .84 12.0 933. Volumes recorded are total accumulation since previous readings. -61-TABLB D (cont*d) RUNOFF AND EFFLUENT LOSSES 1958 - 1959 Date Effluent from Test Area Effluent from "Wall" Area Effluent from Guard&" Runoff" Jan. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 .06 .06 .06 .20 .12 .06 .06 Month total .62 gal. lost 1. 1.5 2.5 1. 2. 2.5 3.5 1.5 1. 1. 3.5 1. 10.5 32.5 gal.. * some losses Volumes recorded are total accumulation since previous readings. 16 11 7 7 26 35 35 26 35 21 35 5.6 20 19 25* 35 26 21 35* 38 35* 75 35* 40 34 43 8 748.6 gal. -62-Date TABLE B (cont'd) RUNOFF AID EFFLUENT LOSSES ' 1958 - 1959 Effluent from Test Area Effluent from Wall Area Effluent from Guards"Runoff" Feb. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Month total Mar. 1 2 Month total .06 .06 .06 .06 .06 .06 .06 .42 gal. .06 706 * some losses 1.2 1.2 1.2 2.3 5.0 2.5 2.5 2.6 2.6 11.0 10. 1.0 3.1 2.1 20. 2.4 1.5 1. 73.2 gal, 2.6 6.0 8.6 gal. 10. 10. 3.2 1.0 23. 13. 9. 32.. 19. 7 . 37. 15. 2.5 4.. 12. * 14. 5.6 9. 2.6 10. 45., 58. 35. 19. 9. 404.9 gal. 37. 2 1 707gal. Volumes recorded are total accumulation since previous readings. -63-xABLB E PIEZOMETER PERFORMANCE 1958 - 1959 November 29 Three piezometers installed near "toe" of s i l o at 12 inches, 24 inches, and floor l e v e l . Tested to ground when immersed i n effluent i n dam approximately equal to 2.3 amps.(high scale) December 1 24 inch c e l l read 0.2 indicating insulation failure 3 12 inch c e l l read 0.6 - surface water leak 4 24 inch, 12 inch c e l l s , both read 0.2 - insul-ation failure 7 A l l wires missing. January 31 New piezometers installed i n test area and near "toe" of silage at 12 inches, 24 inches and floor level. Ground put i n si l o half way between piezometers. Wet reading approximately equal to 2.0 February 0-4Nothing 5 24 inch test area c e l l flooded - surface leakage repacked hole 17 12 inch "toe" c e l l 0.2 - been wet or "leaking" 21 12 inch "toe" c e l l reading down to approximately 0.05 23 Floor c e l l , "toe" - 0.9 - flooded by backed up pool behind dam. 25 Floor c e l l i n test area 0.05 - "leaking" Floor c e l l , "toe" area 0 reading. - 6 4 -APPENDIX F PERMEABILITY EQUATION Tne hydraulic concept of water movement i n a mass uses the p r i n c i p l e of flow as a r e s u l t of differences i n water pressure at d i f f e r e n t points i n the mass. The flow of l i q u i d can be described by the general equation Q, a AV where Q, i s the volume of l i q u i d flowing through a plane surface i n the mass i n a given time; A i s the conducting area and V i s the average v e l -ocity of flow through the conducting area. In a s o i l or simi-l a r mass, neither the conducting area nor the average v e l o c i t y can be e a s i l y computed. Therefore, the general flow equation i s modified to Q=AKi. The t o t a l area of the sample perpendicular to the d i r e c t i o n of flow i s taken as A. The hydraulic gradient i i s taken as the head h divided by the length of t r a v e l , L, through the sample. Then the constant K can be determined experimentally since Q, h, L , and A can be measured. 


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