Bureau of Reclamation Experiences with PVC Seams

 

William R. Morrison & J. Jay Swihart

Bureau of Reclamation, Denver, Colorado 80225. USA

---

ABSTRACT

The Bureau of Reclamation has been using polyvinyl chloride (PVC) plastic in its buried membrane canal lining work for over 20 years. Results of tests conducted on samples of inservice linings indicate that the factory-fabricated seams retain excellent shear and peel strength properties with no apparent signs of deterioration. The practice of using a 1 - m overlap unbonded PVC field seam has proven adequate for most irrigation canal lining applications, but would not be suitable for applications requiring 100% seepage control. Results of laboratory investigations conducted in conjunction with a study on the underwater lining of operating canals with PVC indicate that an adhesive formulated for the repair of vinyl swimming pool liners can be used to make underwater PVC field seams. Results of these investigations also indicate that field seams made in the dry can achieve enough early peel and shear strength development (within 15 min) for placement underwater.
 

INTRODUCTION

Reclamation has used PVC linings for seepage control in irrigation canals for over 20 years. The earliest PVC plastic lining installation was a small experimental section installed in 1957, on the Shoshone Project in Wyoming.1 The first PVC installation under construction specifications (604C-72) was on the Helena Valley Canal, Montana, in 1968. The plastic lining was an alternative to the hot, spray-applied asphalt membrane material.2 (Because the energy crisis in the 1970s caused a significant increase in the cost of petroleum products, coupled with a limited source of supply, the asphalt membrane material was deleted from our specifications.) Over the years, Reclamation has obtained samples of PVC from various installations to determine the aging characteristics of these materials.3 Results of tests conducted on PVC scams from two installations are discussed in this paper.

Laboratory tests were also conducted on PVC seams as part of the research program to develop methods and materials for the underwater lining of operating canals. Reclamation has a number of leaky, unlined irrigation canals that cannot be easily dewatered for lining because of water delivery commitments. Underwater installation of a PVC lining protected with a concrete cover is currently being evaluated.

In addition, PVC seams were evaluated among other seams under a laboratory study Reclamation conducted for the Environmental Protection Agency (EPA) entitled 'Evaluation of Flexible Membrane Liner Seams after Chemical Exposure and Simulated Weathering'.4 The results for the PVC seams are presented in this paper.
 

FIELD PERFORMANCE OF PVC PLASTIC CANAL LINERS

PVC plastic linings were originally used in the rehabilitation of old, unlined canals, especially in areas unsuitable for compacted earth or concrete linings.5 Plastic linings finding wider use in new construction.6.7 The work involves four basic steps: excavation, subgrade preparation, installation of the plastic membrane, and placement of the earth cover (0.3 - 0.5 m in depth) to protect the membrane from the elements and physical damage.

Because of the requirement of an earth cover, membrane linings are restricted to canals having low-velocity flows (0.3 -1 m/s). Also the side slopes should be no steeper than 2.5(H): 1(V) and preferably 3(H): 1(V) to minimize cover stability problems.

PVC is manufactured in roll goods approximately 2 m wide. The roll goods are factory fabricated into sheets wide enough to cover the canal prism and up to several hundred meters in length depending upon its thickness. For most canal lining work, sheets of PVC lining can be joined simply by lapping the downstream end of one sheet of 0.9 m over the upstream end of the adjacent sheet. The PVC plastic has a tendency to adhere to itself and, with the weight of the earth cover, a sufficiently bonded joint is obtained where 100% seepage control is not required. The watertightness of the unbonded field seam is discussed in more detail in the next section. Where a more positive seal is required, the PVC is overlapped a minimum of 0.3 m and a solvent cement (recommended by the manufacturer) applied to a minimum width of 50 mm.

A continuous study is being conducted by Reclamation to evaluate the performance of buried PVC membrane canal linings. Results from two installations-Bugg Lateral, Tucumcari Project, New Mexico, and the Helena Valley Canal, Helena Valley Unit, Montana- are presented.

Bugg Lateral

In the spring of 1961, a small test of 0.25 mm PVC was installed on the Bugg Lateral, Tucumcari Project, New Mexico. The test section was about 228 m in length, and it is the oldest Reclamation installation for which performance data are available for this material. The hydraulic properties of the canal are summarized in Table 1.

TABLE 1
Hydraulic Properties of Plastic-Lined Canals
Protective	Flow	Velocity	Bottom width	Normal water depth	Cover
Canal depth	(m3/s)	(m/s)	(m)	(m)	(m)
Helena Valley	-26	0.64	2.7	28-Jan	0.3
Bugg Lateral	2.66	0.57	4-Feb	4-Jan	0.4
Note: Ratio of side slopes in both canals is 2 (horizontal) to 1 (vertical).

 

Samples were obtained in 1965 (4 years of service), 1970 (9 years of service), 1975 ( 14 years of service). 1980 (19 years of service), and 1988 (after 27 years of service). A photograph taken during the 1980 field sampling is shown in Fig. 1. Results of the sampling indicated the lining was intact below water level, but had suffered some damage from root penetration above the waterline.

 

TABLE 2
Results of Laboratory Tests Conducted on PVC Seam Samples from Bugg Lateral.
	Tucumcari Project, New Mexico
		Typical
Physical	Specification	original	4 Years	9 Years	27 Years
property	requirements	results	of service	of service	of service
Thickness	0.25	0.26	0.26	0.25	0.25
(mm)	10%
Tensile	3	4	4.1	4	5.5
Strength
(kN/m)
Bonded seam	1.95	4	4	3.8	5.8
Strength in
shear (kN/m)
Bonded seam	Not	NDa	ND	ND	3.5
Strength in peel	Required
(kN/m)
aNot determined.

Helena Valley Canal

In the fall and winter of 1968-69, a reach of the Helena Valley Canal, 1930 m in length, was lined with 0.25-mm thick PVC plastic. This was the first PVC lining installation under a Reclamation construction specification (604C-72). The PVC was furnished in sheets 12.8 m wide by 122 m in length. The sheets were accordion folded in both directions for delivery to the job site.

Samples of the lining containing a factory seam were obtained after 9 and 14 years of service. Results of laboratory tests conducted on the factory seam are summarized in Table 3. Test results indicate that as with the Bugg Lateral lining, the factory seams retained their integrity after 14 years of service.


LABORATORY TESTS FOR UNDERWATER LINING OF OPERATING CANALS

Reclamation has been conducting research to develop new technologies for lining canals while they are in operation. The basic concept consists of placing a PVC geomembrane covered with gravel, soil or concrete while the canal remains in operation. The canal would be lined in two or more passes necessitating an underwater field seam in the PVC geomembrane down the centerline of the canal. A 1-m overlapped unbonded seam was planned for this location. As previously mentioned, Reclamation routinely, uses this type of seam (in the transverse direction only) for its PVC-lined canals. Leakage through the unbonded seam was expected to be relatively small since PVC tends to bond slightly to itself under pressure. Seepage measurements obtained for some of these canals, although limited, has supported this expectation. For underwater lining, a study was undertaken to quantify the seepage for this type of seam and to examine the effects of hydraulic head, cover depth and cover material. Additional important information was obtained, quite accidentally, concerning the effect of an irregular subgrade. These results led to a second phase of the study where a new adhesive for bonding PVC under water was examined. Conventional solvents for field seaming in the dry were also examined.

TABLE 3
Results of Laboratory Tests Conducted on PVC Seam Samples from Helena Valley Canal.
Helena Valley Unit, Montana
	Specification
Physical	requirement	Typical	9 years	15 years
property	results	original	of service	of service
Thickness (mm)	0.25	0.27	0.25	0.25
	10%
Tensile strength	3	5.8	5	5.7
(kN/m)
Bonded seam	2.2	5	4.6	5.1
strength in shear
(kN/m)
Bonded seam	Not	NDa	ND	3.7
strength in peel	required
(kN/m)
a Not determined.

 

Phase 1-Unbonded field Seams

The test apparatus for determining seepage through the overlapped seam measures (width by length by height) 1.2 m by 2.4 m by 0.6 m and is shown in Fig. 2. The gravel drain collects the seepage while the geotextile provides a smooth subgrade for the PVC liner. A more representative

subgrade material (i.e. something less permeable than gravel) would obviously reduce seepage; however, an investigation into various subgrade materials was beyond the scope of this study.

Three cover conditions were examined including 25 mm of sand (No. 50 in size), 25 mm of sand plus 50 mm of concrete blocks (200 mm by 600 mm), and 25 mm of sand plus 150 mm of concrete blocks. The voids (approximately 10 mm wide) between the concrete blocks were filled with sand. With the aid of a stand-pipe, tests were run at hydraulic heads of 0.3, 0.9, 1.5 and 2.1 m. Each test was run for a minimum of 24 h to allow stabilization of hydraulic gradients within the gravel drainage layer. Some tests were run for up to 2 weeks to evaluate observed decreases in seepage with time. The results are summarized in Table 4.

Test sets A and B are duplicates with 25-mm sand/50-mm concrete cover and demonstrate the variations seen for identical test conditions. These test sets were meant to approximate the 75 mm of concrete cover. The seepage at 2.1 m of head represents 15-30 liters per day per linear meter of seam and was considered acceptable. A gradual decrease in seepage was seen with time, caused either by fines moving through the overlapped seam and plugging the geotextile and/or gravel drain, or by settlement and compaction of the sand between the concrete blocks.

TABLE 4
See page through Overlapped Unbonded Seam in PVC Geomembrane
Test Set	Cover	Hydraulic head	Seepage
			(m)	(liters m d)
	A	25 mm of sand plus	0.3	0
		50 mm concrete	0.9	0
			1.5	2
2.1	15
	B	24 mm of sand plus	0.3	1
		50 mm concrete	0.9	5
			1.5	-
			2.1	30
	C	25 mm of sand plus	0.3	1
		150 mm concrete	0.9	4
			1.5	5
			2.1	15
	D	25 mm of sand	0.3	15
			0.9	60
			1.5	80
	E	25 mm of sand plus	0.3	60
		50 mm concrete	0.9	400
		(wrinkle in geotextile)	1.5	-

Test set C used 150 mm of concrete blocks rather than the 50 mm used in test sets A and B. No measurable differences in seepage were detected.

Test set D had only the 1.5 mm of sand cover (no concrete blocks) and demonstrated 20 times more seepage than test sets A and B which had 25 mm of sand and 50 mm of concrete cover. This increase in seepage has two causes. The first is the difference in cover load 25 mm versus 75 mm, and the second is the difference in seepage paths. The sand/concrete combination has not only longer but also fewer seepage paths, as the seepage can only occur through the sand between the concrete blocks.

Test set E again had 25 mm of sand plus 50 mm of concrete cover: however, a defect was inadvertently introduced into the subgrade by a fold (wrinkle) in the geotextile. This defect increased seepage by a factor of about 100. As subgrade defects will be impossible to avoid entirely in the field, methods for bonding the seams underwater are needed to assure maximum water conservation.

TABLE 5
PVC Seam Strength Using Special Vinyl Liner Adhesive
	Peel Strength	Shear Strength
Cure condition	(kN/m)	(kN/m)
Air	2.6	10
Underwater	3	12.2
Requirementa	1.8	9.8
a Specification requirements for factory seams

 

Phase II-Bonded field seams

Phase II of the study examined solvents adhesives for field seaming of PVC geomembranes both underwater and in the dry. The biggest challenge was finding a solvent which could be used underwater, as there has been very little experience in this area. Discussions with manufacturers led to the selection of a specially modified bodied tetrahydrofuran solvent used to repair vinyl swimming pool liners.

Test results for PVC seams made both underwater and in air with the special vinyl adhesive are summarized in Table 5. Tests were conducted to determine peel and shear strength after a 24-h cure. Test results indicate that the seams are quite satisfactory and even meet the requirements for factory seams using conventional solvents in the dry.

There was also concern about the rate of seam strength development for the transverse field seams that would be needed every 60 m. These seams would be fabricated in the dry with conventional solvents but then very quickly (perhaps within 15 min) subjected to shear stress as they were placed underwater in the canal prism.

Seam specimens were fabricated in air with a manufacturer-supplied solvent cement and tested for shear and peel strength after cure times ranging from 5 min up to 4 h. The shear strength developed very quickly (within 5 min) and then decreased with time until reaching equilibrium after 1-2 h. Conversely, the peel strength developed rather slowly and required-30- 60 min to develop fully. Shear strength is the more critical as shear is the predominant stress on the seams during installation and service.

Tests were also performed to ensure that specimens, made with conventional solvent and initially cured in air, would continue to cure underwater. Seam specimens were partially cured in air for 5 min and then cured in water for 3 days. These specimens did indeed develop full shear and peel strengths.

RESULTS OF EPA STUDY

In 1986, Reclamation completed a study for EPA entitled 'Evaluation of Flexible Membrane Liner Seams After Chemical Exposure and Simulated Weathering'. In this study, 37 geomembrane seams, both factory and field were evaluated. The PVC seams included in the study are listed in Table 6.

The seams were subjected to six chemical solutions, brine and water immersion, freeze/thaw cycling, wet/dry cycling, heat aging, and accelerated outdoor aging for periods of up to 1 year. Effects of these environmental conditions were evaluated using shear and peel strength tests before and after exposure. The tests were performed under dynamic load at room temperature and under static dead load at 50° C.

The rate of grip separation for both the peel and shear tests was kept the same (50 mm/min) to determine if there was any direct correlation between the two properties. Also, a 25-mm wide test specimen was used in both tests.

Results of tests conducted on seams subjected to water immersion, freeze/thaw cycling, wet/dry cycling, and heat aging are summarized in Table 7. These environmental conditions are those often encountered in Reclamation's hydraulic applications.

TABLE 6
Type of PVC Seams Evaluated in EPA Study
	Sample	Type of	Manufacturer	Seaming	Seam
	No	Seam	Fabricator	Method	Width
	1	Factory	A,C	Solvent adhesive	25
	2	Factory	A,D	Thermal-dielectric	19
	3	Field	Aa	Solvent adhesive	50
	4	Filed	Ab	Solvent adhesive	88
	5	Filed	Bc	Solvent adhesive	75
a Solvent adhesive furnished by fabricator C.
b Solvent adhesive furnished by fabricator D.
c Solvent adhesive furnished by manufacturer B.

Test results indicated that except for heat aging, the samples performed satisfactorily with very little change occurring to either the shear or peel strength. The heat aging samples exhibited stiffening due to plasticizer loss from the material.

Of the two factory seaming methods used for the PVC, the seams made with the solvent adhesive exhibited higher shear strength, whereas those made dielectrically produced higher peel strength values. The higher shear strength was primarily due to the wider factory seam for the solvent adhesive seam. All failures occurred in the parent material, except for the peel tests on the PVC solvent adhesive seam, where the failure occurred within the seam itself. No appreciable difference was noted in the performance of the two seaming methods however.

TABLE 7
Results of EPA Study on PVC Seams
		Sample 1		Sample 2		Sample 3		Sample 4		Sample 5
	Test Condition	Shear	Peel	Shear	Peel	Shear	Peel	Shear	Peel	Shear	Peel
Original		10	2.6	9.3	6.7	8.3	2.7	9.3	3.4	10.4	4.3 Water immersion at
23°C	6 months	9.9	2.8	9.1	6.6	8.5	3.1	10.5	3.8	10.6	4.3
	12 months	10	2.9	9.3	7.3	9	2.7	10.7	3.9	11.7	4.4
Heat aging at 60°C	4 weeks 9.2	3.1	9.3	6.5	8.8	3.8	10.4	4.2	10	4
	8 weeks	9.1	3.1	9.2	6.9	9.4	3.2	9.3	4.3	11.3	5.1
	13 weeks	10	3.3	9.5	6.9	9.1	4.1	10	4.4	11.4	3.9
Freeze/thaw at:	10 cycles	10.1	2.7	8.9	6.7	8.4	2.4	9.6	4	10.5	4.4
	20 cycles	9.7	2.8	9.3	7.3	9.3	3	10	4	12.4	4.7
	50 cycles	10.1	2.8	8.4	7.2	8.9	2.7	9.1	3.8	10.4	4.7
Wet/dry at:	10 cycles	10.2	2.7	9.4	6.9	8.4	2.6	10.3	3.8	11.2	6
	20 cycles	9.9	2.7	8.5	6.5	8.8	3.3	9.5	3.7	10.2	3.2
	50 cycles	9.9	2.7	10.1	6.7	8.2	2	10.6	4.3	10.7	4.6
Note: Test values are expressed as kN/m width of seam. One freeze/thaw cycle consisted of freezing for 16 h at 6.7?C and thawing for 8 h in room temperature water. One wet/dry cycle consisted of 16 h water immersion followed by 8 h of drying at 37.7?C.

CONCLUSIONS

The Bureau of Reclamation has been using PVC in its buried membrane canal lining work for over 20 years. Results of tests conducted on samples of inservice linings indicate that the factory seams retain excellent shear and peel strength properties with no apparent signs of deterioration.

A 1-m overlap, unbonded PVC field seam appears to be adequate for most irrigation canal lining applications, but would not be suitable for landfills or hazardous waste installation where 100% seepage control is required.

Results of laboratory tests also indicate that the solvent-bonded field seams can achieve early peel and shear strength development which is advantageous for underwater lining applications.

Laboratory tests conducted on an adhesive sealant formulated for the repair of vinyl swimming pool liners indicated that it can be used to make underwater PVC field seams.

Results of laboratory tests involving various environmental aging conditions indicate that there is no appreciable difference in the performance of solvent or dielectrically made factory seams.
 

REFERENCES

1. Hickey, M. E. Investigations of plastic films for canal linings. Research Report No. 19. A Water Resources Technical Publication. Bureau of Reclamation, Denver, Colorado, 1969.

2. Geier, F. H. & Morrison, W. R. Buried asphalt membrane canal lining, Research Report No. 12. A Water Resources Technical Publication. Bureau of Reclamation, Denver, Colorado, 1968.

3. Morrison, W. R. & Starbuck, J. G. Performance of plastic canal linings. Bureau of Reclamation Report No. REC-ERC-84-1. Denver, Colorado, 1984.

4. Morrison. W. R. & Parkhill, L. O. Evaluation of flexible membrane liner seams after chemical exposure and simulated weathering, US Environmental Protection Agency. Report No. EPA/600/S2-87/0l5, Cincinnati, Ohio, 1987.

5. Wilkinson, R. W. Plastic lining on the Riverton Irrigation Project. Proc. ASCE Irrigation and Drainage Speciailty Conference. Flagstaff, Arizona, 1984.

6. Starbuck, J. G. & Morrison, W. R. Flexible membrane for closed basin conveyance channel. San Luis Valley Project, Colorado, Proc. International Conference on Geomembranes. Denver, Colorado, 1984.

7. Weimer, N. F. Use of polyvinyl chloride liners for large irrigation canals in Alberta. Canadian Geotechnical Journal. 24 (1987) (2) 252-9.
 

---

 

REC-ERC-84-1

PERFORMANCE OF PLASTIC CANAL LININGS

January 1984

Engineering and Research Center

U. S. Department of the Interior
Bureau of Reclamation

 

Table18. -Physical properties test results for PVC membrane linings on Bugg Lateral.
Tucumcari Project, New Mexico, installed spring 196l.- Continued
		Sample	Sample	Sample
	Specifications	No. B-6764	No. B-7022	No. B-7023
Physical property	requirements	(14 years	(19 years	(I 9 years
		of service,	of service,	of service,
		BWL)	BWL2 )	AWL1)
Thickness, mm (mils)	0.25 (10)	0.24 (9.6)	0.24 (9.6)	0.21 (8.2)
percent change	±10	-14.3	-14.3	-26.8
Tensile strength, N/mm (lbf/in)	3.0 (1 7)	4.2 (24.2) L3	4.6 (26.4) L	5.0 (28.6) L
		4.6 (26.1) T4	5.2 (29.8) T	4.7 (26.9) T
percent change		-5.5 L	+3.1 L	+1 1.7 L
		+14.5 T	+30.7 T	+18.0 T
Elongation percent	225*	268 L	211 L	151 L
		274 T	188 T	188 T
percent change		-35.0 L	-48.9 L	-63.3 L
		-40.7 T	-59.3 T	-59.3 T
Modulus at 100 percent	Not required	2.4 (13.8) L	3.6 (20.5) L	4.6 (26.0) L
elongation, N/mm (lbf/in)		2.4 (13.8) T	4.2 (23.9) T	4.2 (23.9) T
percent change		+21.1 L	+79.8 L	+128.1 L
		+34.0 T	+132.0 T	+132.0 T
Elmendorf Tear, grams	1500*	3000 L	3000 L	450 L
		2865 T	2200 T	1300 T
percent change		+63.9 L	+63.9 L	-75.4 L
		+25.1 T	-3.9 T	-43.2 T
Impact resistance	Not more than 2	5 tested	Not	Not
	specimens out		determined	determined
	of 10 shall fail	5 failures
	at -18ºC (0ºF)
Plasticizer content, percent	Not required	34.1	27	21.6
percent change		-14.3	-32.2	-45.7
Bonded seam strength,	65	Not	Not	Not
percent of parent material		determined	determined	determined
1 AWL denotes above normal waterline
2 BWL denotes below normal waterline
3 L denotes longitudinal direction
4 T denotes transverse direction
* Minimum, each direction

 

 

Canal data:		Established
		seepage
b = 5.00 ft, d =4.40 ft, s:s =1.5:1, and		rate
WP =20.86 ft.		(7.4806 gal/ft3) (2169 ft2)
Length =104.25 ft.
Average seepage rate for section is 8.2		= 0.0005 (ft3/ft2)/d
gal/d.		Reach 5A, station 599+00, PVC lined:
Seepage section has a wetted surface of
(104)(2y0.86) = 2169 ft2.		Canal data:

 

For more information call 800-OK-LINER today!