Bentonite Clay vs. Geomembrane Pond Liners: Seepage Control, Installation, Cost & Durability

Below 500 gallons per acre per day (0.47 mm a day), use an HDPE membrane or a bentonite GCL. Between 1,000 and 5,000 gallons per acre per day (0.94 to 4.7 mm a day), with the right soil, a soil-bentonite liner does the job for less.

The 1.5 mm HDPE sheet alone costs $3.8 to $12.9 a square metre. A GCL costs $5.4 to $16.1. Soil-bentonite with its earthwork lands at $18.4 to $20.2, and both synthetic routes install for $5.4 to $53.8.^[1]

HDPE holds about 150 years at 25°C. Push the temperature to 33°C and that falls to roughly 70 years. A GCL loses an order of magnitude of permeability once calcium in the water passes 500 mg/L.^[2][3]

172 mm a year at the 500 limit, 5.7% of a 3 m water column, and 1.7 million litres off a 10,000 m² pond. An acre at the same rate still loses 1.9 m³ a day.

Conversion factors. 1 gallon = 3.785 litres, 1 acre = 4,046.86 m² = 43,560 ft², 1 ft² = 0.0929 m², 1 lb/ft² = 4.88 kg/m², 1 ft = 0.3048 m, 1 mil = 0.0254 mm. Leak figures assume a 10,000 m² pond. Summer evaporation is taken at 6 to 13 mm a day, so swap in your own weather data.

Send the pond area, the depth and a water analysis. We come back with a membrane thickness, a roll width and a quantity, and every roll is batch tested to GRI-GM13 before it ships. Request a quote

Seepage control

Compacted clay has to reach 1×10⁻⁷ cm/s. US EPA and China’s GB 16889 use the same figure. China’s sanitary landfill code, GB 50869-2013, caps a GCL at 5.0×10⁻⁹ cm/s, twenty times tighter than the clay limit.^[5][6]

A GCL arrives 6 to 10 mm thick and replaces a clay layer of 610 mm, so you trade 1/61 to 1/102 of the thickness for a smaller permeability number. That trade holds up only when the carrier geotextiles and the overburden pressure do their part. The saving shows up fastest on sites that would otherwise truck in compacted clay and compact it in lifts.

Field surveys across 108 Chinese landfills found 15 defects per hectare and 122 cm² of defect area per hectare, or one hole every 667 square metres. Strict quality control in the US and Europe gets it down to 3.^[7] Counting the holes on your own project takes one pass of an electrical leak location survey before the cover soil goes on.

Brazil inspected 21 landfill projects over 15 years, 95 hectares in all, and repaired 194 holes. Call it 2.1 per hectare. Punctures and heavy machinery caused 83% of them, and 43% ran longer than 10 cm.^[8]

An older survey of 89 projects over 265.2 hectares found 4 to 22 leaks per hectare. A second one covered 682 hectares and pulled out 4,761 holes, an average of 6.98.^[8]

Now put a 10 cm hole on the books. Take free orifice flow with a discharge coefficient of 0.6 and 0.3 m of head, and the theoretical ceiling sits near 990 m³ a day. Scale that 43% share and you are looking at about 0.9 holes of that size per hectare.

Materials and structure

The bentonite loading is nothing alike. The GCL carries 0.75 to 1.13 lb/ft², or 3.7 to 5.5 kg/m², against 29.3 kg/m² in a 12-inch soil-bentonite liner. That works out to five to eight times the material.

Weight on site changes the whole working day. The GCL order runs to about 48 rolls and lands at roughly 48 tonnes. A 1.5 mm HDPE liner weighs about 14.1 tonnes, or 1.41 kg per square metre, and it shows up as a welding job rather than a materials handling job.

What the codes require

Installation

Thickness gets quoted in mils, and the conversion between mils and millimetres matters the moment a Chinese mill spec meets a US datasheet. 30 mil is 0.76 mm, 60 mil is 1.52 mm, 80 mil is 2.03 mm, 120 mil is 3.05 mm.

A 7 m roll over 10,000 m² leaves about 1,430 m of seam. Drop to a 5 m roll and the welding runs to 2,000 m. Same pond, 570 m of extra work, and every metre of it needs hot wedge welding and a seam test.

Overlaps eat material. The 150 mm side lap takes 3.3% of a 4.6 m roll width. The 500 mm end lap takes 1.1% of the 45.7 m roll length. Add edge trim and your waste budget is 4% to 5%.

Cover soil runs 3,048 m³ of fill over 10,000 m² at 300 mm. At the $4.39 a cubic yard used in the NRCS cost sheet, that comes to about $17,500.

Cost

Rules for reading these numbers. Market ranges come from published quotes, so treat them as orders of magnitude and leave out tax, third-party testing and contingency. Factory bulk and retail membrane prices differ by 3 to 8 times, so keep the two apart.

The low end of the installed range covers a normal pond, the high end covers a landfill-grade system with clay, drainage and leachate collection. The percentage split comes from a different sample than the unit prices, so do not subtract one from the other.

With cover soil in the mix, earthwork and cover reach $14,166.53, or 17.3% of $81,711.50. Without cover soil, earthwork alone is $7,081.07, or 9.5%. Two different lines from one cost sheet, so do not add the 9.5% to the cover figure.

Bentonite is $66,539.71 of the total, or 81% to 89% of the job. Check the quantity against the spread rates in NRCS 521C. Silt takes 0.375 lb/ft² per inch of liner, silty sand takes 0.5 and clean sand takes 0.625. The New Jersey figure of 6.0 lb/ft² is exactly the silty sand rate times 12 inches.

The membrane split looks different. Materials take 45% to 60% of it. Laying and welding take 25% to 35%. Subgrade work takes 10% to 15% and testing takes 5% to 10%. On 10,000 m² the installed range runs $54,000 to $538,000, and the layer-by-layer breakdown explains why the spread runs that wide. In practice thickness, roll size and freight move the number more than the resin price does.

How long it lasts

Antioxidants are the clock inside HDPE. Soaked in water at 25°C they are spent in about 40 years. Under a saturated sand layer with 260 kPa of load it takes about 120 years. Cool the site to 20°C and the same test stretches to 200 years.^[2]

Run the three-stage model and a landfill primary liner holds about 150 years at 25°C or below. A secondary liner runs past 300 years at 15°C and about 400 years at 10°C. At 33°C the primary liner drops to roughly 70 years.^[2]

A separate dataset puts 20°C with steady oxygen past 300 years and 40°C past 45 years, a gap of at least 6.7 times.^[2] Ground temperature and UV exposure belong in the specification, not in the afterthoughts.

GCL performance turns on water chemistry. 0.0125 M calcium chloride works out to 1,387 mg/L, of which 501 mg/L is calcium. 0.6 M sodium chloride is 35 g/L.^[3]

Either one lifts permeability by an order of magnitude or more, and the long test runs needed 30 months and 686 pore volumes to settle down. Site liquids get tested the same way, with the GCL permeated by the actual leachate or process water.^[11]

The 10,000 square metre bill

Over 150 years the membrane material costs $253 to $860 a year on 10,000 m². Over the 30-year design life most engineers work to, it runs $1,267 to $4,300 a year. Quote both numbers, because the 150-year figure hides repair and patch work.

Tell us the pond size, the depth and the side slopes. We return the roll count, the tonnes on site and the seam length to weld, so the figures above become your own project numbers instead of a range.
Send your pond dimensions

Choosing between them

• A clean water pond up to 8 ft (2.44 m) deep, with silt or silty sand on site, takes a soil-bentonite liner from 6 inches (152 mm) up. This is the case most farm reservoir liners fall into.
• Where you want factory quality, a rough subgrade and a tight schedule, a bentonite GCL fits, with 150 mm side laps, 500 mm end laps, 300 mm of cover and 4% to 5% put aside for waste.
• Where calcium runs above 500 mg/L, or the water is salty or hard, HDPE membrane carries the job. A GCL in that liquid loses an order of magnitude of permeability.
• High head, machine traffic, long exposure and a water level that keeps moving all point to HDPE membrane. Pick 7 m rolls to save 570 m of welding, and go to a textured sheet where the side slopes are steep.
• A regulated waste site needs a composite liner, membrane over GCL or compacted clay, 60 mil (1.52 mm) and up, with leachate held under 30 cm. Hazardous waste goes further, because the rule there calls for two liners with detection between them, and a landfill bottom liner usually adds a drainage layer and a second membrane.
• Earthwork is only 9.5% of the soil-bentonite bill, so trimming the subgrade saves very little.

Common mistakes

• Treating a GCL as a thin clay layer. Bentonite content is one fifth to one eighth of a 12-inch soil-bentonite liner and thickness is 1/61 to 1/102 of a clay layer, so it leans hard on confinement.
• Judging a membrane by its thickness. Hole density drives the leak rate, and Chinese landfills average one hole every 667 m², five times the rate achieved under strict control in the US and Europe.
• Buying calcium bentonite to save money. NRCS 521C calls for sodium bentonite with a free swell of at least 22 mL/2g, and calcium bentonite does not get there.^[12]
• Skipping cover soil. NRCS 521 requires PVC and GCL to sit under at least 12 inches (305 mm) of soil, with the largest particle at 3/8 inch and 1/2 inch.^[13]
• Assuming the job ends when the GCL is down. Punctures and machinery cause 83% of the holes, mostly while cover soil goes on, and a dry lap or an unseated slope turns into a leak path. On membrane jobs the anchor trench is where that failure usually starts.
• Ignoring temperature. Eight degrees of difference in service temperature takes the estimated life down by more than half.

A short checklist

1. Fix the leak limit first. 500 gallons per acre per day is 0.47 mm a day, and that is the line between a soil treatment and a synthetic liner.
2. Test calcium in the water. Above 500 mg/L, drop the bentonite route for a membrane.
3. Set the thickness. Soil-bentonite goes 6 to 24 inches (152 to 610 mm) with depth, and HDPE takes 1.5 mm as the floor.
4. Set the subgrade tolerance. GCL needs bumps under 12 mm and ruts under 25 mm, and the membrane needs a geotextile cushion.
5. Write the joining and testing plan. Laps and bentonite for the GCL with 4% to 5% waste, air pressure and destructive samples for the membrane.
6. Plan the cover. 10,000 m² takes 3,048 m³ of fill, about $17,500.
7. Cap the largest particle in cover soil at 3/8 inch (9.5 mm) and 1/2 inch (12.7 mm).
8. Testing runs 5% to 10% of the installed cost, and an electrical leak survey before and after cover pays for itself.

Questions people ask

Can a GCL sit under water for years

Yes. China’s GB 50869-2013 uses it under an HDPE membrane in a composite liner with a permeability cap of 5.0×10⁻⁹ cm/s, and NRCS 521C asks for protection against desiccation and water level movement.

Can I skip the membrane

Below 500 gallons per acre per day, NRCS lists bentonite GCL alongside HDPE, LLDPE and EPDM as synthetic liners worth considering, and the main pond liner types all sit in that same list. Once calcium passes 500 mg/L, add the membrane on top.

How thick should the membrane be

NRCS 521 splits it at 60 mil (1.52 mm) for wastewater and 30 mil (0.76 mm) for clean water. China’s GB 50869-2013 asks for 1.5 mm minimum, and 2.0 mm where waste reaches 20 to 50 m high.

GRI GM13 covers 0.75 to 3.0 mm. A 60 mil sheet must reach a density of 0.940 g/cc, carbon black of 2.0% to 3.0%, and either 100 minutes of standard OIT or 400 minutes of high-pressure OIT.^[14] Those four lines are the ones to check on a submittal before you approve it.

Which one costs less

On 10,000 m² the material bill is $38,000 to $129,000 for 1.5 mm membrane, $54,000 to $161,000 for GCL, and $164,000 for the bentonite in a soil-bentonite liner. Once installed, both routes land inside $5.4 to $53.8 a square metre, and the difference is who gets the money.

How do I find a leak

Run a dipole survey over the covered liner. ASTM D7007 covers the method, and once you have the location you patch the panel or swap it out.^[15]

Can I use both together

US EPA 40 CFR 258.40 and 257.70 require a composite liner with a flexible membrane over at least two feet of compacted soil and HDPE at 60 mil minimum. China’s GB 50869-2013 writes the HDPE-plus-GCL build-up into the code as well.^[5][16]

How long does each one last

Ground temperature sets the membrane bracket, about 150 years at 25°C and about 70 years at 33°C. Calcium in the water sets whether a GCL works at all, and above 500 mg/L you change the design.