Anaerobic Reactors

An OnLine Primer

Need to know more? Send us an email or feel free to search our online series with our give away search engine:



To a certain extent, perhaps the following graph may be one of the most helpful charts ever to help assess the various anaerobic  technologies/configurations available for targeting a specific wastewater.

SRT = total active biomass concentration / net growth of active biomass

The first task is to get reasonably accurate measures of waste flows and COD/solids loads and other critical parameters such as TKN and sulfates. Then one can work toward process selection and reactor sizing. Other treatment objectives such as nitrogen removal need to be factored into the process selection and sizing. Without treatability testing, it usually is not beneficial to use complex modeling so a simple, educated spreadsheet approach to reactor sizing and estimating of biogas production is OK for starters.  



Abattoirs are well suited targets to low-rate, anaerobic process because of the usually low COD and high O&G levels.  It may be also possible to design as competitive, higher rate systems.  As with most every anaerobic approach, process temperature is key.  One will want to avoid seeing an anaerobic treatment plant for wastewaters that contain grease, such as meat processing, milk, cheese, etc., operate at less than 32C. If client or engineers  insist, they have to deal with the consequences of designing and operating at lower temperatures. First , a larger reactors. Second, most of the grease will float to the top and form a scum layer that in some cases with meat slaughtering operations has approached six ft (2 m) in depth. This scum layer is very difficult to breakup especially if the reactor is covered with a membrane type material. These problems may occur at 32C, but to a much lesser extent.  Therefore stick to and maintain the higher temperature range.  Below 20°C removals are simply settling, hardly any anaerobic biological process contribution.

Abattoir wastewater from fowl is particularly difficult to deal with in that it contains a lot of protein and gut waste which is high in nitrogen. The protein is slow to degrade with strong odor potential, as rotting protein tends to have. Cadaverine for example is one of the organic chemicals produced by decaying flesh that gives it the ugly smell. In warm climates, picking things like open, trickling filters or aerobic biotowers where biomass and flesh could get trapped on the media and wherein the decay odors could get passed directly to air leaving the tower, would be definitely unadvisable. Many of abattoirs in the U.S. are using oversized and relatively inexpensive oxidation ditches to process these wastewaters. 

Clarifiers usually are not good options for pretreatment of slaughter-house wastes because of the stringy characteristics, the fats, and the odors. Grinders followed by screens and DAF or centrifuges usually are better.


Anaerobic treatment is the best option for both citrus and wet milling wastewaters.  To a certain extent high rate configurations such as UASBs and EGSBs can be considered as well as low rate reactors.  Low rate rate systems will feature design loading rates probably somewhere in the 0.5 to 1.0 kg/m3/day range, so as with high rate designs one can roughly work out the approximate volumes.  Low rate systems are good for wastewaters such as thin stillage.

Corn wet milling wastewaters are ideal candidates for anaerobic treatment. Whole stillage is best treated using completely mixed reactors; thin stillage -- after solids removal -- can be treated using high rate processes.  One can give preliminary reactor sizing if the general wastewater characteristics are known -- COD, TSS, VSS, TKN, etc. -- and flow rates.

In general one requires as much information as possible about the wastewater characteristics including COD, TSS, VSS, alkalinity, pH, VFAs, and temperature in the reactor. Also, a schematic diagram of the reactor would be helpful. Lemon processing wastewaters often contain limonene, which is an organic solvent and can cause damage to anaerobic granules. Any information on limonene concentration in the wastewater will be helpful.


HAFs are not good candidates for treating dairy wastes, Because of the high FOG and high biomass yields, the media tends to plug and float. UASBs are much better if the FOG is less than about 100 mg/L. Low rate or completely mixed reactors usually are best overall.



Landfill leachates are difficult to treat anaerobically. While HAFs have been used, their history is not good. The high inorganic dissolved solids usually include substantial amounts of calcium which precipitates as calcium carbonate in the reactor. This can eventually plug the media.

Landfill leachates contain significant amounts of nonbiodegradable or very slowly biodegradable organics and often contain substantial amounts of color. Heavy metals usually are a minor problem since they will be absorbed by the biomass. Nitrogen levels usually are high so that ammonia released or there will be a demand for oxygen to satisfy nitrification. High O&G figures probably point to other hexane  extractable materials --- probably organic acids. It is highly recommended highly that treatability tests be conducted so that the designer will know exactly what efficiencies can be achieved. 

A possible approach could be a low rate mixed reactor or a contact process that includes a mixed digester and a clarifier for solids removal. Since the wastes usually are not amenable to granulation and because of the high dissolved salt content, landfill leachates usually are not treated using UASB or EGSB reactors. A low rate type reactor would be designed for operation at up to 1 kg/m3/d and a contact process would be designed to operate at around 4 kg/m3/day. Still be careful with this one. The problems can be difficult to manage.


In general, one needs to know more about this wastewater before making a decision about the best way to treat it. At first thought it may not seem a good candidate say for TF treatment because of the high strength and the consequent need for high recycle rates, in addition to the concern for odors. Anaerobic would be much better, but pharmaceutical wastewaters often contain antibiotics and sulfates that can make anaerobic treatment difficult.  Because of the nature of the waste, a treatability test should be commissioned to review the actual characteristics and determine if there are any constituents that would cause problems with anaerobic treatment. 


Biomass from sugar wastewaters form granules readily, so there is little risk to using UASB/EGSB reactors. For this wastewater type UASB/EGSB processes are much better than say attached growth, anaerobic processes such as HAFs.  Sugar wastewaters produce a lot of biomass that can accumulate in the media and cause floatation and damage to the reactor. Unfortunately one has seen this happen too many times, in some cases within the first year.  There are some ways to avoid these problems, such as installing gas purge systems to blow the excess solids out, but not every project/design makes provisions for this   HAFs are much better for acetic acid and protein wastewaters that contain little suspended solids.

UASB or EGSB reactors are best suited for treating bottling wastewaters but low rate installations do exist. One advantage of the low rate reactor is that it is much more forgiving and requires less EQ volume up front.  One would expect the cost is not greater than an UASB or IC reactor; otherwise low rate alternates would not be promoted.


The greatest challenge in breweries is control of toxic agents used as lubricants and biocides in brewing operations, and diatomaceous earth can cause significant abrasion of granules.


Craft breweries generally would be too small to justify on-site anaerobic treatment so would have arrangements to discharge their wastewater to the city's wastewater treatment plant. Some craft breweries may be required to separate grain solids before discharging liquid wastes. On-site AD not only would require the anaerobic treatment unit but also would require gas handling and burning equipment, which usually rules out AD at small sites.

AD will not treat the waste to a high enough degree to discharge to a stream. If you are going to discharge it could be better to use a small scale sequencing batch reactor which is one of many activated sludge process variants.


Feedlot wastes are treatable anaerobically, but require a process configuration that allows the owner to process high solids content.


There is a lot of activity among all major suppliers to treat food wastes. A number of alternatives are marketed heavily. The engineering challenge is to match the waste characteristics with the best process configuration. Also, most of the suppliers treat their designs as proprietary information.


The Germans have a lot of experience with manure digestion using large mixed reactors (we would hesitate to say completely mixed), and some high-solids digesters (usually plug-flow). Manure is digestible but management of mixing and sludge removal/disposal is of high importance.

Lagoons are not good options for treating manure or grease because of inability to mix and remove waste solids. So they fill up quickly and if covers are used, the grease floats and accumulates beneath the cover with little or no digestion. Nevertheless, some industries use lagoons for treating slaughterhouse wastes throughout the world with frequent cleanout and cover replacement.



Rough solids digester volume can be preliminary sized estimating COD as three times the grease value and about 1.1x for non-grease solids and about 2-3 kgCOD/day per m3 (OK for mixed and plug-flow digesters; 0.5 to 1.0 for engineered lagoons, but < 0.5 for basic covered lagoons).

One can compare this preliminary/front-desk sizing with say designs using a solids loading of about 0.08 lb volatile solids/cu.ft./day (~ 1.3 kg/m3/day).



This parameter usually does not affect anaerobic treatability unless it is above 95%, in which case the solids would be 5% or 50,000 mg/L TSS. If the moisture content is less than 50%, the solids may be too heavy to mix.

This parameter represents salts or grit. High salt content (TDS > 15,000 mg/L) can inhibit methane production. Manure can have high ash content due to dirt and mud. Grease wastes usually are low in ash.

Most volatiles of concern with anaerobic treatment are the volatile organics such as gasoline, hexane, toluene, etc. These compounds usually are not toxic but also are not biodegradable in anaerobic reactors, but can cause dispersion of solids.

Grease has a COD/weight ratio around 3:1, so grease presents a high COD loading potential and at temperature below 30C usually do not degrade efficiently and accumulate in anaerobic reactors as floating grease balls or rise to the surface of the digester and forms scum layers. This is especially a problem with covered anaerobic lagoons.

Manures usually have low calorific content because of the straw, hair, and dirt. At high loading rates, the ammonia nitrogen can be sufficiently high to inhibit methane production. We usually try to keep ammonia-N below 1500 mg/L for safe operation.


One thing we recommend strongly is that you should NOT put waste activated sludge back into the anaerobic digester. This tends to fill the digester with inert solids and over time will fill the lagoon with inert solids or dilute the activity of the anaerobic biomass.


Water hyacinths can be anaerobically digested. A lot of work was done on this topic at the University of Florida in the 1980 - 90s by Anne Wilkie and other researchers. The problem is that the methane gas yield is low, averaging 25 to 42% of theoretical, thereby leaving 58 to 75% as residual biosolids or soluble COD. The higher methane yields seem to be associated with digestion of dried and ground water hyacinths, which adds processing costs. It is very unlikely that anaerobic digestion of water hyacinths will be economical based on the value of the methane gas but it might be economical if treated by co-digestion with municipal sludge simply as a way to reduce their biomass. In any case, the economic viability depends on the cost of methane gas.


Tannery wastewaters are very difficult to treat anaerobically because of the salts, acids and chemicals used for processing the hides.  One further  needs to know the type of tanning process -- vegetable, chrome, etc. -- and the general characteristics -- COD, VSS etc. If the wastewater is treatable, one would expect one will need to use DAF to remove the solids before anaerobic treatment. A low rate type reactor could be OK, but probably design COD loading rate should be close to 1.0 kg/m3/day. 


An upflow velocity of say 6 m/hour is quite high for a UASB that does not already contain granular biomass. If the digester sludge you use for seeding is not granular, you will need to reduce the upflow velocity, probably to less than 1 m/hr, until granulation occurs. This could take from one to six months. You will need to feed a readily biodegradable COD (sugars, alcohols, benzoic acid, etc. but not acetic acid) during this time to support the growth of granules. Then increase the upflow velocity slowly until stable operating conditions are achieved. The only short-cut [we know] is to start the UASB with granular biomass from another UASB reactor.


One possible alternative to deal with hydrogen sulfide  is to add ferric or ferrous chloride to the influent to the anaerobic reactor to tie up the sulfide as ferrous sulfide. This alternative works well but will increase the density of the sludge so may not be best for UASB or EGSB reactors. It takes about 1.1 kg of Fe per kg of sulfide if no competing reactions are present.  Another alternative is to strip the sulfide from the gas stream using a caustic scrubber at pH > 9.  This method would require a spray tower, a caustic (NaOH or KOH) feeder, and a sump for recycling the scrubber water. This method works well but can require substantial amounts of caustic because some carbon dioxide also will be scrubbed as sodium or potassium bicarbonate.  Biofilters work well for vent gases but would be very large and expensive for biogas streams. The Hiperion and similar processes provided by a number of vendors convert the hydrogen sulfide to elemental sulfur, but can be quite expensive. Activated carbon works well but probably will be very expensive.


For the purpose of rough calculations we can, whatever the units used, and using say COD as constituent, combine these two formulas:

Yn = Yo * ( 1 + 0.2 * Kd * SRT) / ( 1 + 1.2 * Kd * SRT)

Yn = bugs in tank * 100 / ( SRT * kgCOD/day * COD removal points)

Thus we have
bugs in tank * 100 / ( SRT * kgCOD/day * COD removal points) = Yo * ( 1 + 0.2 * Kd * SRT) / ( 1 + 1.2 * Kd * SRT)

Solving for one of the SRT we get the coveted iteration formula:

SRT = bugs in tank * 100 * ( 1 + 1.2 * Kd * SRT) / ( kgCOD/day * COD removal points * Yo * ( 1 + 0.2 * Kd * SRT) )

From now on it's just a matter of allowing Excel to perform circular reference iterations (you may have to enable this setting in your Excel version).

For those unfamiliar with iterative methods (a numerical analysis classic) one way to find a root of an equation is to solve for the variable and give iterative methods a try.
Were we to find a root for say
a * x^2 + b * x = (a * x + b) * x = 0

One possible iteration formula to try could be solving for x as follows:

x = 1 / (a * x + b)

While it may seem laughable to do it for this case the truth is that it works greats for the not so immediate cases. With minor tweaks it can be used for UASBs, HAFs, BVFs, BNR TFs of all sorts.


Equalization tanks can be in line or side-line, but it does not matter for anaerobic process design. We believe in-line is better because it reduces odors and in general recommend only submerged mixing. The air only causes oxidation of organics and release of odors. If odors are a concern, then the EQ tank needs to be covered and vented to an odor control unit.

One thing we've found with earthen basins, is that the sloped walls greatly restrict the effect of submerged mixers and cause solids to accumulate at the edges of the basin. These solids also make it difficult to clean the edges of the basins and can create odors and allow insects and weeds to grow. In general we recommend adding vertical concrete walls at the water line for the earthen basins.


The inlet and RANS laterals should be placed on blocks -- concrete should be OK -- to which the laterals should be fastened. They should be heavy enough to avoid flotation of the pipe. Vertical concrete walls around the top of the basin are used to attach the cover.

Does this type of BVF anaerobic reactors use a double membrane gas holder?

That's your decision. There are several options for lagoon covers. Some are double layer; others are single. We personally prefer the type of cover that floats on the surface with gas collection at the sides and ends. This type of cover avoids wind damage. We have actual plant records showings some problems with covers that allow gas collection under the cover.

The gas is channeled to the perimeter via lateral floats. The floats create a high point and the rain trench creates a low point. A micro gradient is formed between the two structures. Rainwater flows downward into the trench and gas bubbles formed under the cover travel along the gradient to the high point where they are pulled to perimeter piping via negative pressure. The perimeter piping terminates at a draw-off pipe (preferably through the slope and not the cover) to a flare or genset.



Need Some Help? - James C. Young Environmental - Balestie, Irwin & Balestie