When Could You Possibly Operate An Anaerobic Digester Without Mixing?

BACKGROUND
It has been shown that after secondary treatment, anaerobic digestion is tied for the second highest consumption of electrical power at an average water resource recovery facility (WRRF; SAIC, 2006). Despite that, there has been limited research on how to optimize digester mixing. Many design engineers and regulators take the approach that it is better to err on the side of excess mixing power. The implicit assumption is that excessive mixing has little or no detrimental effect. That may not be a good assumption. In a landmark study on digester foaming, Miot et al (2013) showed that reducing the fraction of time the digester mixing system was in operation in an egg-shaped digester reduced foaming and variability in the daily biogas flow without reducing the total biogas flow. Eventually the best results were achieved with the mixing system turned completely off. There is also much anecdotal evidence that at least some digesters function well without external mixing. Occasionally, digester mixing systems break down and the error is only discovered much later often without impacting digester performance. In other cases, owners and operators may not have the budget available to fix a mixing system that fails unexpectedly or prematurely and are forced to operate the digester without mixing, sometimes for years or even decades, without obvious signs of loss of performance. Collectively these cases raise the question of whether anaerobic digesters can be operated without external mixing, relying only on the natural mixing provided by the biogas generated in the digester (and mixing from the heat loop) or if it is only possible under certain conditions.
METHOD
Digestate in a WRRF is a non-Newtonian, shear-thinning fluid, in that its apparent viscosity reduces at higher shear rates (Taylor et al., 2020). As the shear rate is increased from near zero, there is initially a sharp drop in apparent viscosity, but eventually there is little change with increasing shear rate. One theory for optimizing digester mixing is to operate the digester at a point where there is a limited response to changes in shear rate without approaching the point of diminishing returns (Krugel & Stallman, 2018). Presumably operating under these conditions would limit the variability of viscosity throughout the digester. The authors suggested two distinct rates of response (slopes in the curve) that might be targeted for successful operation: -0.33 and -0.50 Pa.s2, as shown in Figure 1. The general applicability of these targets remains to be confirmed experimentally. Rheological testing of a sludge sample at different solids concentrations allows a series of curves, such as shown in Figure 2, to be developed. This means that the shear rate and the corresponding minimum mixing intensity can be determined at each different concentration. Figure 3 shows the minimum mixing intensity for three different plants as a function of digester solids concentration. Note that the rheological testing at Plant A used two different cup and bob sizes in the rotational viscometer, and those produced slightly different results, confirming the limitations of current test protocols. Both Plants A and B digest primary sludge only, but their measured viscosities are quite different: at around 4% total solids, the estimated minimum mixing power for both plants are similar, but it is obvious that at higher concentrations the requirement for Plant A increases much faster than for Plant B. Plant C, digesting only waste activated sludge (WAS), has much higher sludge viscosity and requires much more mixing intensity. This confirms that rheological properties vary significantly from between WRRFs, so that it is necessary to test the rheology at each WRRF. Pre-treatment, such as thermal hydrolysis, may be used to reduce sludge viscosity. Pretorius et al. (2018) showed how to calculate the mixing power provided by the biogas flow in digesters. The procedure can be reversed, to estimate the minimum gas flow required to provide the mixing intensity calculated from sludge rheology. The gas flow can be used to estimate the minimum volatile solids loading rate required, based on an assumed volatile solids reduction (VSR) percentage and a gas yield (m3/kg VSR or cf/lb VSR). This is not a new approach: the USEPA (1976) stated that if a volatile solids loading rate of 6.4 kg/m3.d (0.4 lb/cf.d) could be maintained, natural mixing would be sufficient. Depending on feed sludge concentration and volatile fraction, that loading rate would translate into an HRT of approximately 6.5 days, confirming that this is an aggressive loading rate. The VSR depends on the hydraulic retention time (HRT) in a once-through digester. The HRT, in turn, is directly related to the volatile solids loading rate. An alternative way to present the conditions required for natural mixing to suffice is as a maximum HRT.
RESULTS
Of the three plants evaluated, only the Plant A data yielded viable results for relying on natural mixing. The Plant B data, determined for a recuperative thickening project, started at a relatively high solids concentration of 3.8% where a viable solution would require a specific digester design, such as a deep digester or a high aspect ratio. For Plant C the viscosity throughout the tested range (2.9% and above) was too high to yield a viable solution. Figure 4 shows the minimum volatile solids loading rate required for natural mixing as a function of digester solids concentration, based on the rheological data presented in Figure 2. The figure suggests that up to a solids concentration of approximately 2.5%, natural mixing should suffice without excessive volatile solids loading rates. The figure also shows that for this particular sludge, the criterion recommended by the USEPA (1976) of 6.4 kg/m3.d (0.4 lb/cf.d) is quite conservative and would allow natural mixing to suffice up to a solids concentration of almost 4%. Figure 5 shows that maximum allowable HRT for the same rheological properties. The figure suggest that a target HRT of 15 days can be maintained up to a concentration of approximately 3%. The full manuscript will include a discussion of the impact of SWD and digester shape on the mixing requirements. Other factors, including grit accumulation and blending of the feed into the digester contents will also be addressed.
CONCLUSION
A method is presented that allows a comparison to be made between the mixing provided by biogas generation in the digester and the mixing required to ensure a relatively even digester viscosity to be maintained. The results show that site-specific sludge viscosity determines whether natural mixing by itself could be sufficient. The method would allow significant power savings at WRRFs to be realized. Eliminating or significantly reducing the digester mixing power would allow WRRFs to make significant progress towards achieving energy neutrality or positivity.