Demand-Driven Gas Production Using Co-Substrates from the Dairy Industry

INTRODUCTION Water resource recovery facilities (WRRF) have become more and more efficient within their system limits over time. With increasing degree of energy efficiency, further optimization measures are becoming more challenging, especially if the WRRF is understood as an autarkic energy system. Multiple WRRF have already achieved a positive energy balance. Further optimization possibilities consist of smoothing load profiles in order to avoid feeding electricity into the public grid or actively taking part on energy markets. The main option to actively manage loads on WRRF is the combined heat and power plant including the gas storage. It offers good opportunities to decouple energy consumption and energy production, especially since the effects on the waste water treatment performance of the WRRF remains largely unaffected. Nevertheless, gas storage capacities are limited. Especially with increasing size of the WRRF, the storage capacity is only sufficient for a few hours. Fig 1 shows the ratio of gas storage capacity and daily biogas yield for WRRF with different sizes. Energy rich co-substrates with high biological degradability can compensate small gas storage capacities if they are dosed with foresight. In this study different feeding strategies were evaluated and their effects on process stability and dewaterability. MATERIAL AND METHODS Semi-technical investigations were carried out in four identical digesters (37 °C) with a capacity of 20 liters. The feed consisted of municipal sewage sludge (mixture of primary (PS) and surplus sludge, SS) and flotation sludge (FS) from the dairy industry. All digesters were operated with a hydraulic retention time (HRT) of 19.5 d over a period of 2 months. 41 days were evaluated within this study. Fig. 2 shows the experimental set up of the 4 reactors. D4 was solely fed with sewage sludge continuously with an interval of 2 hours. D1 was continuously fed with sewage sludge and additionally once a day with a charge of flotation sludge. D3 was also fed continuously with sewage sludge, flotation sludge was given twice a week. D4 was fed intermittently with sewage sludge and flotation sludge. Operating parameters can be taken from Tab.1. Analysis were conducted according to the analytical methods specified by the German Institute for Standardization. Additionally, dewatering tests were conducted with a filter press, including sludge conditioning with polymers (Kemira) with an amount of 10 mg active substance per g TS. RESULTS and DISCUSSION The share of flotation sludge on the total feed was around 20 w-% for D1 to D3 over the observation time. As the COD content was twice as high in the co-substrate (95,100 mg/l compared to 52,500 mg/l in the raw sludge) and its high degradability the share of the flotation sludge on the total methane gas produced amounted to 35 - 40 %. In D1 gas production rate can be elevated by 81 % within 6 hours or even 112 % within 2 hours (cf. Fig. 2). In D2 gas production rates were elevated by 80 % within 20 hours or 130 % within 4 hours. Maximum rates were reached after 1 - 2 hours after feeding. In Fig. 4 a COD balance over a period of 41 days is displayed. The balance shows the measured values of the COD loads. The degraded COD (COD-deg) was determined by the difference between the COD in the feed (COD-input) and the COD in the output stream (COD-output). To cross check the results, the biogas content was also converted into a COD-load (COD-biogas) using the conversion factor of 0.35 m ³ CH4/kg COD ). D1 and D2 showed the same degradation rates despite having different organic loading rates (OLR) per feeding events (1.5 and 4 g COD/(m ³ d)) due to different feeding intervals and flotation sludge amounts per feed. Overall degradation rate was 66 % respectively 69 % for both reactors. Specific methane production was almost identical. D3 showed a higher specific methane production. Thus, the dry matter content increased in this reactor steadily with this kind of feeding pattern. All reactors were fed with the same moderate loading rate related to the entire observation time. It is known, that famine phases have a major impact on the biocenosis and can stimulate biogas yield which could be an explanation for the higher specific methane production in D3. All digesters were stable regarding the volatile fatty acids (VFA) concentration. Fig. 3 is showing the course of organic acids after feeding the flotation sludge. Depending on the amount of flotation sludge fed a steep increase in the concentration is observed and depending on the amount of flotation sludge fed the maximum gas production rate is reached. Most importantly all reactors reached low concentration after 6 to 12 hours prior to the next flotation sludge feed. Dewatering test showed similar values for D1 and D2, cf. Fig. 5. D3 showed significant lower dewatering properties. CONCLUSION Flotation sludge have a high energy density which is a multiple of the energy density of biogas. It can be used in order to compensate small gas storage capacities when used with foresight. They are able to increase gas production rates within a relatively short time after feeding. Maximum turnover rates are reached for highly degradable substrates after 0.5 - 1.5 hour. Implementing unknown substrates afford a tight monitoring at the beginning to understand kinetics regarding VFA accumulation. Highly degradable substrates will show steep increase in VFA for a defined time. Accordingly, minimum feeding intervals can be defined in order to avoid VFA accumulation. Dewaterability will detoriate using co-substrates especially protein and lipid rich substrates as used in this case. Thus, gains from the additional biogas yield must be weighed against the losses from sludge disposal. Nevertheless, as long as sludge is fed continuously feeding intervals seem to have a minor impact on the dewaterability properties of the digested sludge. In the following the results are going to be set into context of a full scale plant by evaluating data from a WRRF of 350.000 PE (cf. Fig. 6 and Fig. 7).