Bench-Scale Biomethane Production Tests
Coagulants - Alum increased potential specific CH4 methane yield by 15%
Kemira C557 1.8 mls in 1,000 mls barn flush water, detention time 40 minutes
400 mls solids, 73 minutes 275 mls solids, Kemira C573 and C581 less effective
respectively. 2 mls of alum (35.74 Degrees Baume, 48.7% Sluminum Sulfate Solution,
MC Chemicals) in 500 mls of barn flush, floc volume 150 mls, detention time 7 min.
Williams Farm Due Diligence Summary
A due diligence investigation was performed by Remtech Engineers to determine if dairy manure from the Williams farm would have sufficient biogas methane concentrations and flowrates to justify a biogas renewable natural gas project. A rate of 130 cfm with over 50% methane was generally required. This farm served as a baseline example to build a data base for viable vendors for agricultural anaerobic digesters and associated equipment and dairy digester operation and training.
Everett Williams (2017 Georgia Farmer of the Year) owns a 1,700 lactating cow dairy farm located Morgan County, south of Madison, Georgia (1141 Broughton Road Madison, GA 30650). An initial visit was made on May 23, 2020, to document farm operating procedures, sand recycling, manure flushing, manure pond cleanouts, manure field applications, and his interest in participating in a renewable energy project. A site drawing was prepared and updated throughout this investigation.
On June 14, 2020, samples were collected from manure flush influents to the Weep Wall and Castle (Weep Wall) basins and one Castle sludge sample. Samples were collected during cow cooling water misting that may not have been representative of recycled water flush operations. Samples were prepared and sent to the University of Georgia's Agricultural lab for analysis.
Screen tests were performed on recycled sand bedding to determine sizing of conventional sand/manure separation equipment. Results confirmed sand to be typical of concrete sand (79.69% retained on 5 to 60 mesh and 53.33% retained on 5 to 40 mesh). Manure flush samples were found to have low total solids (1.16%, desirable range 4 to 7%) and total volatile solids to total solids ratio (31.7%, desirable range 70 to 80%+) and low Biological Oxygen Demand/Chemical Oxygen Demand (BOD/COD ratio – 0.262, less than 0.3 indicates potential toxics present). One sludge sample was collected from the Castle Pond appeared to have higher solids and total volatile solids present.
Reducing manure flush water volumes/frequencies or changing to manure scraping to increase solids concentrations was not a viable option according to Mr. Williams.
Coagulant tests were performed with several flocculants including alum from C & S Chemical and Kemira C-577 that were successful in separating solids and liquids. Other coagulants tested included ferric chloride, calcium chloride, powdered activated carbon, Kemira C573, Kemira C581 that were not as efficient. Solids concentrations with flocculants increased to 1.8 - 6.2% but total volatile solids/total solids (TVS/TS) ratios ranged from 10.8 to 38.17% that were too low. Separated liquids also had low TVS/TS ratios ranging from 41 to 50%. Floc particles in thickened solids were found to be fragile for and may pose issues with thickened sludge removal from settling or dissolved aeration systems.
Spoiled corned silage (estimated 600 tons/yr.) was also tested as a co-digestate and found to have a TVS/TS ratio of 30.9.
Mr. Williams employs sand lanes followed by gravity “weep wall” solids removal (used by less than 1% of farmers and is considered an advanced technology by the State of California). Weep walls removes and concentrates higher percent solids than conventional mechanical systems were sampled and found to be 15% within 2 ft of the surface (literature reports up to 33%). A decision was made to investigate harvesting sludge from these ponds as a feed source to anaerobic digesters. Sludge samples were extracted from both the Castle and Weep ponds on July 22, 2022, by deploying Remtech's 20 ft sampling trailer and dragging a jon boat across the hard crust (1 ft deep) of the Castle Pond and pulling sludge samples from the Weep Pond banks (with a 1.5 ft hard crust). Total pond sludge solids samples diluted with 23 to 33% fresh barn flush ranged from 9.1 to 11.9% and TVS/TS ratios increased to acceptable ranges from 62.6 to 79.41%. Low BOD/COD ratios ranged from 0.09 to 0.10 (below 0.3 indicating potential toxins present). Potential toxics include recycled disinfectants from parlor operations (Heavy Duty CIP Detergent, Acidoxy-San, Pensep, and Alphacide Activator), feed antibiotic additive Rumensin, and a manure odor control agent “More than Manure”. Rumensin is reported to reduce methane production rates by up to 40% and possibly increase methane bacteria acclimatation rates up to more than several months. Typical Rumensin feed addition rates are 29 mg/kg feed. Mr. Williams Rumensin feed rate was calculated to be 9.49 mg/kg feed. The methane rate limiting Rumensin concentration is not known.
Remtech uses proprietary bench-scale anaerobic digestor reactors that simulate full-scale digestors using gas flowrate and composition instrumentation used on full-scale digestors. Methane concentrations and production rates were found to range from 50 to over 70% at flowrates ranging from 39.76 to 93.15 cfm. Methane production rates began within 1 to 3 days after starting reactors confirming that sludges were already in an active anaerobic digestion state. Reactors were set up in one 3 liter, three 2-gallon and one 5-gallon reactor with gas concentrations measured with an RKI Eagle 2 infrared monitor (0 to 100% methane, % oxygen, and 0 to 60% CO2). Theoretical methane production rates were based on 242 mls of CH4/mg of TVS added ranged from 63 to 98 cfm CH4. Methane production rates from 8 operating anaerobic digesters scaled to the Williams herd size ranged from 43.56 to 125 cfm CH4 (DVOs plug flow digesters).
Methane production flowrates were based on 2,000,000 gallons of sludge harvested from both ponds every 6 weeks as indicated by Mr. Williams on visits leading up to August 25, 2020. On August 25, 2022, Mr. Williams indicated that both ponds had not been cleaned since April of 2022 through August 25, 2020, or 4 months (16 weeks) that put the sludge feed stock volume in question. In addition, it was discovered that Ponds 1, 2, and 3 were cleaned out by passing waste from these ponds through the Weep wall pond that would potentially interrupt feed to digesters for a period ranging from 6 to 10 weeks.
Sludge temperatures were found to increase from ambient 78oF to 100oF in the Weep Pond and 78 to 90oF in the Castle Pond. Insitu sludge heat generation would potentially reduce sludge heating requirements to raise temperatures to 95oF. Wastewater from the parlor was found to have a pH of 7.3 and chlorine concentrations of 0.4 to 0.5 ppm, and other analytes that may pose a potential toxic impact on methanogenic bacteria.
Methane production flowrates were initially overestimated by adding the reactor head space to the volume collected in 1-liter Tedlar bags per elapsed time. Bubble calibrator gas flowrates were conducted next that produced more accurate rates. It should be noted that reactor pressures were only 0.37” H20 (measured with a Dwyer magnehillic gage) that may have limited the volume of gas transmitted to Tedlar bags and bubble calibrator. Another method was to measure the exhaust gas volume from the RTI meter that yielded a volume correction factor of 1.6 to 2. With an average flowrate from AD bench-scale reactors of 50.11 cfm (eliminating the high and low measurements), flowrates are potentially too low (80.17 to 100.22). Two readings were over 90 (93.15, 100.89), 1 reading over 70 (78.48), and the remaining 11 readings were between 39.03 and 54.57 cfm CH4.
This farm is a potential future renewable energy project if sludge volumes are verified and/or increased with Mr. Williams potential farm expansion, addition of co-digestates from other farms or high energy waste feedstocks, elimination of potential toxins, confirmation of methane production flowrates by resampling fresher manure sludges, and retesting sludges in bench-scale digesters. Note that the reference data scaled down from the patented DVO plug flow reactor suggested biogas flowrates of 125 to 135 cfm (considerably higher than other reactors) should be confirmed. Contacting owners of operating dairy manure reactors and confirming methane production rates and profitability would help in selecting the type of anaerobic reactor. Using the table below @ 0.1 cfm/cow @ 60% methane = 102 cfm.
The Sustain RGN AD reactor (employing the patented Train\AD system - currently under construction 20 miles from Mr. Williams farm) is connected to three farms J W Farms (2 farms) and Sunrise Farms near Eatonton, GA – estimated dairy herd 2400+. This is the first of its kind in the US and consists of screening, electrocoagulation, dissolved aeration (CO2 inert gas), inground fixed film barrier in a multidirectional chamber, filtration, and return of liquids to the front end of the process.
Methane production rates - Methane Produced = 242.7 mls CH4/gm of TVS added. Potential methane yields can also be estimated per animal type in cfm/animal units. These Biogas production rates can be off + 100% on full scale operations. Lab generated Biogas production rates based on Co-digestion substate additions of whey and FOG are referenced below:
Bench-Scale Biomethane Production Tests
Remtech’s proprietary bench-scale anaerobic digester reactors simulate full-scale digesters using gas
flowrate and gas composition instrumentation used on full-scale digesters. Manure sludge samples
from two weep wall thickeners from a 1,700 cow sand lane/flush manure system located in central,
Georgia were run in four (4) digesters.
TVS, TS, TVS/TS ratios, BOD/COD Ratios - Biological Oxygen Demand/Chemical Oxygen Demand
(BOD/COD) ratios less than 0.3 indicates potential toxics or insufficient organics are present. Total
volatile solids to total solids ratio ranged from 70 to 80%+.
Methane concentrations and biogas production rates were found to range from 50 to over 70% at
flowrates ranging from 39.76 to 93.15 cfm. Methane production rates began within 1 to 3 days after
starting reactors confirming that sludges were already in an active anaerobic digestion state.
Reactors were set up in one (1) 3 liter, three (3) 2-gallon and one (1) 5-gallon reactor with gas
concentrations measured with an RKI Eagle 2 infrared monitor (0 to 100% methane, % oxygen,
and 0 to 60% CO2) that measures methane concentrations without H2S interference. Theoretical
methane production rates were based on 242 mls of CH4/mg of TVS added ranged from 63 to
98 cfm CH4. Methane production rates from 8 operating anaerobic digesters scaled to a 1700 dairy
herd size ranged from 43.56 to 125 cfm CH4 (DVOs plug flow digesters).
Sludge temperatures were found to increase from ambient 78oF to 100oF in the Weep Pond and
78 to 90oF in the Castle Pond (type of Weep Pond). Insitu sludge heat generation would potentially
reduce sludge heating requirements to raise temperatures to 95oF. Wastewater from the parlor
(containing disinfectants) was found to have a pH of 7.3 and chlorine concentrations of 0.4 to 0.5 ppm,
and other analytes that may pose a potential toxic impact on methanogenic bacteria.
Methane production flowrates were based on 2,000,000 gallons of sludge harvested from both ponds every 6 weeks as indicated by the farmer on visits leading up to August 25, 2022. Each reactor was placed in heated baths with temperatures set at 95oF. Biogas was collected in one liter Tedlar bags. Daily gas production rates were recorded by the volume of gas collected in bags. Incidental flowrates were measured with a bubble flow calibrator.
Gal Castle AD Sludge Reactor
2 Gal reactor with 1 liter Tedlar Bag (3.3 liters - 3.3 lbs S4 Sludge, 3.3 lbs S5 Sludge, 1.1 liter Barn 5 Flush, 95oF Temp Bath). 4.4 liters total reactor sludge volume, gas generation for 25 days
pH adjusted with HCL from 7.57 to 6.825 and mixed, gas production next day @ 65.5% methane
At day 3, oxygen concentration rose to 13% and methane production fell off. pH adjusted again at day 7 from 7.194 to 7.05 with HCL - this additional drop in pH appeared to shock methanogens reducing methane concentration and biogas production rates.
Took 8 days to get oxygen levels below 2% when methane production, picked up.
Methane concentrations were less than 50% throughout most of the test period.
During test period opened lid, mixed 4 times, and reinjected with CO2/nitrogen.
Avg Flowrate 37.05 cfm CH4
2 Gal Weep AD Sludge Reactor
2 Gal reactor with 1 liter Tedlar Bag (3.3 liters sludge, 3.3 lbs of S4, 3.3 lbs S5, 1.6 liter of Barn 5 Flush, 95oF Temp Bath), Thicker than Castle that appears to be pumpable. 4.9 liters total reactor sludge volume, gas generation for 36 days, with methane concentrations over 60% for most of the test period.
pH adjusted from 7.2-6.835 with HCL.
During test period opened lid, mixed 1 time, and reinjected with CO2/nitrogen.
Avg Flowrate 49.98 cfm CH4
2 Gal Composite Sludge Reactor
(0.86 liters of S4 & S5, S-2, S-3 sludge each and 250 mls of S-1 and 330 mls S-2 Crust, 1.32 liters Barn F5 Flush, 95oF Temp Bath), weight of sludge in reactor 16.37 lbs. 5.31 liters total reactor sludge volume, gas generation for 31 days
pH was not adjusted at 7.32
No mixing during test period
Methane concentrations over 60% for most of the test period.
Avg Flowrate 49.99 cfm CH4
5 Gal Composite Sludge Reactor
1.5 gal of S3 and 1.5 gal S5 and 0.6 liters Barn 5F Flush less 280 ml sample pulled, 13.35 liters, wt 28.4 lbs, 7-9% solids. 13.35 liters total reactor sludge volume, gas generation for 22 days
CO2/N gas injection mixed 2 times during test period, formed pin holes in 2” mat formed on top of reactor. Methane concentrations less than 50% during 9 of the test days when O2 concentrations exceeded 1%
Avg Flowrate 38.46 cfm CH4
158,000/2,000 = 118,500/1,500; 39,500 cfm/500 cows biogas
1,700 cows = 134,300 cf/day or 93.26 cfm, tests 50 cfm
Digester Estimated Costs-
Average AD manure biogas yields of 20 to 30 cm/wet metric ton, is the minimum to just meet the required amount of biogas that is needed to offset capital costs and provide profitability (DEA 1995). Costs of installing a digester varies widely depending on the size, type, and operation. Typical capital costs for AD digester systems ranges from $1,000 to $2,000 per cow (or $1,186 to $2,373/cow adjusted for inflation in 2022 dollars). Operating AD costs, are approximately $588,000 per year for a 2,000-cow dairy, or $294 per cow. (Source: 2018 Livestock and Poultry Environmental Learning Community article).
The CDFA (California Department of Food and Agriculture Dairy Digester Program) funded 117 dairy farm projects between 2015 and 2021. All but one of them were covered lagoon digesters (see the fact sheet by Hamilton for a discussion of digester types). Those funded in 2015 generated electricity, but most of those funded in 2017 and later have renewable natural gas (RNG) as the end-use of the biogas. The AgSTAR database lists the herd sizes for 19 of these digester projects that are producing RNG. For those projects, the average herd size was 7,479 cows and the average project cost was $947/cow or $4,384,479 total before deducting the CDFA grant. These projected costs do not include off-gas treatment systems for conversion to RGN or sludge heating systems.
For the 1,700 cattle project estimated in 2022 in Central Georgia, the off-gas treatment system was $3.5 MM or $2,058/cow for a 250 cfm system (oversized for an estimated 150 cfm system) plus $2,373/cow for AD system or ($4,431/cow) $7.53 MM. Annual O & M costs at 10% of capital costs are estimated at $753,300/yr. Assuming there is a linear scale up factor for a 22,000 cow heard the treatment and AD system is estimated at $97.5 MM plus annual operating costs of $9.7MM.
An estimate projected by DH in November of 2022 - $80 MM, $5 MM/yr O & M. Off-gas treatment and O & M projected costs can have a significant impact on project estimates. In addition, system repair costs due to undersizing, corrosion impacts, sludge heating costs, air pollution controls on heating systems, variations in feed energy inputs, toxic materials (odor control agents, feed antibiotics, disinfectants, design errors, and operator errors), can also impact system estimates.
Estimated RNG project costs for a 2,000-head dairy would be on the order of $8 million dollars ($4,000/cow, adjusted for inflation $4,745/cow) for on-site digestion and $6 million for centralized digestion 9, resulting in RNG production costs of $28-$35/MMBtu (2019, Renewable Natural Gas Project Economics).