Optimization of mixing in a Chinese dome digester for tropical regions

The growing demand for energy because of population growth, lack of clean energy and inadequate availability of natural resources have led to the growing demand of anaerobic digestion technologies in rural areas of many developing countries. Biogas a renewable, clean and efficient source of alternative energy which can be used as a substitute for fuels such as firewood, charcoal and cattle dung, used by rural dwellers. The domestic digester is an example of an anaerobic digester usually applied in a single decentralized system mostly in rural areas of developing countries and can serve as energy solution to meet their cooking needs. Among all household digesters, the Chinese dome digester (CDD) is the most popular and most implemented reactor because of its reliability, requirement of low maintenance and long lifespan (Chapter 2).

Mixing in the Chinese dome digester (CDD) depends on the change of slurry level in the digester and extension chamber during gas use and could be regarded as intermittent natural mixing. Mixing is an important process in anaerobic digestion for establishing contact between micro- organisms and feed, for homogenization of temperature throughout the digester, and preventing settling and floating layers. However, mixing is limited in the CDD and are therefore operated   at long hydraulic retention times (> 40 days) and low influent total solid (TS) concentrations (≤7%) when compared to forced mixed reactors (intermittently or continuously), leading to a large reactor volume and higher cost (Chapter 2). In this thesis, mixing was optimized in the Chinese dome digester without the inclusion of moveable parts with lower water dilutions (high influent TS, 15%) at reduced hydraulic retention times 40 and 30 days.

First, Chapter 3 examined the effect of higher volumetric gas production on mixing during the anaerobic digestion of cow manure in Chinese dome digesters (CDD) at ambient temperature (27-32º C) in comparison with mechanically mixed and unmixed digesters at laboratory scale. Six digesters (two of each type) were operated at two different influent total solids (TS) concentration ranges, viz. 3-7.3% and 6-15 %, at a hydraulic retention time (HRT) of 30 days for 319 days. The impeller mixed reactors were mixed at 55 rpm, 10mins/hour, the unmixed digesters were  not mixed and the Chinese dome digesters were mixed once a day releasing the build-up gas pressure. Significant differences were observed among the three types of digesters at both influent TS concentration applied in this study. The impeller mixed digesters exhibited better biogas production and treatment efficiency, followed by the Chinese dome digesters (hydraulic mixed) and the unmixed digesters. The stirred reactor operated between 3-7.3 % TS concentration produced 20% more methane than the Chinese dome and 37% more methane than the unstirred digesters respectively at steady state conditions. At applying double influent TS concentrations, the reactors showed lower specific biogas production and higher VFA concentrations with few exceptions. The VFA accumulation was more pronounced in the unstirred digesters and Chinese dome digesters. It could also be seen from the results that double TS concentration did not produce better reactor performance (based on specific methane production and VFA concentrations) in the CDDs despite higher volumetric biogas production rate. The natural mixing induced by biogas production did not yield sufficient mixing. In addition, the hydraulic variation or mixing cycle in Chinese dome digesters may not suffice for the treatment of cow manure at TS concentration of 10% and above.

In Chapter 4, the Residence Time Distribution (RTD) technique was  applied  to  evaluate mixing of liquid and solid phases in laboratory scale Chinese dome digesters. Appropriate tracers were studied over a theoretical hydraulic retention time (HRT) of 30 days in the three different digesters (impeller mixed, unmixed  and  CDDs)  each  at  influent  concentration  of ca. 7.5 and 15 % TS concentrations. The impeller mixed reactors had the lowest dead zones followed by the CDDS and lastly the unmixed. The reactor performance in terms of methane production was consistent with the evaluation of the RTD results. The reactor type and mixing modes had direct impact on reactor hydrodynamic and eventually reactor performance. At both TS concentrations, the hydraulic reactors had considerable dead zones because the mixing viz. hydraulic variation is inadequate. The CDD (hydraulic) digester therefore needed optimization for improved hydraulic variation to achieve optimized mixing cycles without use of moveable parts or external energy.

A model for prevention of biogas emission from inlet and outlet of the Chinese dome digester was developed and validated with pilot experimental data in Chapter 5. The model predicted well the reactor pressure (PG) and the slurry displacement in the expansion chamber, inlet pipe (h) and inside the digester (hG). A better decision on the location and heights of theoutlet, expansion chamber and inlet pipes can be made using the approximation model and will therefore prevent emissions during the zero biogas consumption periods.

In chapter 6 the optimized CDD with self-agitating mechanism with the  inclusion  of  two baffles was investigated at a pilot scale (digester volume =500 L) and compared with the conventional CDD (as blank) at 15% influent TS concentration at two HRTs (30 and 40 days). The reactors were operated at ambient temperature, 27- 33 ° C. The optimized digester showed better digestion efficiency and process stability, while the blank was unstable throughout the study period. The optimized Chinese dome digester has a self-mixing or agitation cycle of two minutes using the produced gas without a moving part. The optimized digester showed superior digestion treatment efficiency, and more stable in terms of VFA (mainly acetate) concentrations than the conventional reactor. The improved Chinese dome digester with baffles showed better performance than conventional design and the reactor can be operated at  high  influent TS (15%) concentration. Therefore, this implies a smaller reactor volume could be achieved at high loading rate at reduced HRT (< 40 days), and eventually reduction in reactor cost.

Lastly, a multiphase computational fluid dynamics (CFD) was applied to evaluate the improved mixing in the Chinese dome digester at a pilot scale in Chapter 7. The digesters studied in Chapter 6, are one conventional CDD and the other, improved (baffled) CDD. The optimized digester under goes self-agitating cycles created by the pressure variation from the produced biogas with the aid of a baffle, whereas the blank does not self-agitate. The length of the baffle in the optimized digester was estimated using the model developed in Chapter 5.

The CFD models were validated using the pressure and biogas experimental data from the pilot study of same digesters. The phases and flow fields were analyzed in relation to mixing and then digestion performance in both reactors. The necessary intermittent mixing performed by the produced gas in the optimized digester achieved without moving parts, improved mixing and  the hydraulic characteristics of the optimized digester. The lift force created when the produced biogas flow below the baffle from the compartment A to B and the change of flow direction of the slurry in the digester, inlet pipe and expansion chamber resulted in improved mixing of the of reactor content sufficient to improve digestion performance.