A case study for a cost-benefit-based, stepwise optimization of thermo-chemical WAS pre-treatment for anaerobic digestion

All experiments were performed with sludge of a conventional activated sludge process from the WWTP Zirl (Tyrol, Austria), a municipal WWTP treating wastewater from 61,000 population equivalents. The samples were taken after polymer dosage and mechanical excess sludge dewatering with a final content of total solids (TS) from 4 to 6% (w/w), volatile solids (VS) of 70% and a pH of 7.

Total solids (TS) and volatile solids (VS) were both measured gravimetrically according to standard procedures (EN 12889, 15169) and the pH was determined using a pH-probe (744 pH Meter, Metrohm, Switzerland). All samples were measured in triplicates.

Experiments were made in triplicates at temperatures of 39, 50, 70, 85, 100, 120, 150, 180 and 200 °C with a reaction time of 1 h, a time that had been found optimal by a number of studies [19]. Ambient pressure disintegration up to 100 °C was performed in 50 mL tubes heated in a water bath. To ensure a reaction time of at least 1 h at the given temperature, the core temperature of one sample was monitored. The high pressure disintegration from 120 to 200 °C was performed in a microwave heated pressure capsule (Speedwave, Berghof^®) with the same sludge amount, heating and holding time. For the subsequent determination of disintegration parameters, the liquid and solid fractions were separated by two consecutive centrifugation steps at 3200×g and 37,000×g for 10 min each. The CODs after consecutive centrifugation was slightly higher than conventional CODs estimation by 0.45 µm filtration (<10%). However, this method was preferred to avoid intense filter plugging. The effects of the degree of disintegration were negligible (<1%).

To determine the most suitable PT temperature, several parameters related to subsequent biogas production were measured to increase the reliability of the results:

COD_S was measured photometrically in the supernatants after appropriate dilution with deionised water (Macherey–Nagel Nanocolor^® CSB 1500). K was measured by atomic absorption spectrometry [21] after dilution with 1% CsCl as matrix modifier (Analytik Jena, ContrAA 700), and DOC was determined using an elemental analyser (Shimadzu, TOC-V CPN). For DOC and K, respective values for the dilution liquids were determined and subtracted from the results of the samples.

Finally, a highly disintegrated reference was generated by mixing the sludge 1:1 with 1 M NaOH and heating it to 90 °C for 10 min [22]. The sludge treated in such a manner was considered to have 100% and the native sludge 0% disintegration level and the corresponding COD data were called COD_max and COD_S0, respectively. Using these parameters, an additional value for the evaluation of the degree of disintegration (DD) [5] was calculated in order to compare the achieved data to literature data following Eq. (1)

After the evaluation of treatment temperatures, the most promising temperature was applied for all further experiments: In triplicates, 50 mL of WAS was mixed with a suitable amount of 1 M NaOH to achieve concentrations of 0.08, 0.04 and 0.02 mol L⁻¹ and was immediately heated to 70 °C for 1 h in the water bath under continuous stirring. For the alkaline reagents ash and aluminate, one representative molarity was selected (0.04 M) to treat the sludge in the same manner. To do so, the alkalinity of both substrates was determined by titration of the diluted product with 0.5 M HCl to pH 4.3, and the results were used to produce a 1 M stock solution to be used for the experiments. The ash (bottom ash with a small amount of fly ash) was obtained from the biomass power plant Bioenergie Kufstein (Tyrol, Austria) and was sieved to 2–5 mm particle size prior to use, the aluminate was achieved from Donau Powerfloc flow. The pH of any chemically pre-treated sludge was determined after reaching the final temperature using a pH-probe (744 pH Meter, Metrohm, Switzerland).

The most promising combination of chemical and thermal pre-treatment regarding enhanced expected Biogaspotential due to DD_COD in combination with low energy input was used to determine the biogas potential in relation to an untreated WAS sample, using the Automated Methane Potential System (AMPTS; Bioprocess Control, Lund, Sweden) and following a standardised procedure [23]. Digested sludge from the WWTP Zirl was used as seeding sludge and allowed to degas at 37 °C for 7 days. In triplicates, negative controls with seeding sludge only, positive controls with microcrystalline cellulose (Alfa Aesar A17730) with a theoretical biogas potential of 740–750 L kg _VS ⁻¹ and mixtures of pre-treated and native samples with a ratio of VS_Sample/VS_{Seeding sludge} = 0.5 were anaerobically digested at 37 °C for 21 days.

TS, VS, pH, total proteins (DC Protein-Assay, Bio-Rad, Hercules, US) and the total COD were measured before and after the fermentation.

In order to estimate the dewaterability of the pretreated sludge in comparison to untreated sludge, two digestion runs were performed, one with untreated sludge, and the second one with pretreated sludge. 1 L batch reactors, containing 0.5 L of sludge were fed daily with pretreated or native substrates for 83 days (digester load: 2 kg_VS m³ day⁻¹). Dewaterability was analyzed thermogravimetrically by the consulting company Kläranlagen Beratung Kopp, Lengede, Germany and expressed determining the value TR_A [24], a parameter defining the achievable TS of the sludge and the required amount of polyelectrolyte (PE) (ZETAG^® 8160 BASF). This method was found to be the only suitable lab-scale method to predict actual dewatering results in full scale units [24].

Applying a cost-benefit calculation, the advantages gained by the usage of the optimally pre-treated sludge were evaluated, including the actual costs for energy, NaOH, polyelectrolyte (PE) as well as digestate disposal costs and subtracting the benefit of the savings of electrical power deduced from biogas production. To account for heat energy, no additional costs were assumed for the pre-treatment of WAS, since heat at 70 °C is sufficiently available at the WWTP Zirl. The costs for NaOH were calculated assuming a prize of 350 € t⁻¹, digestate disposal was accounted for with 70 € t⁻¹, the cost for the polymeric flocking agent with 3 € kg⁻¹ and the savings for electricity were calculated with 0.10 € kWh⁻¹ (average purchase costs for major customers). For service and maintenance, personnel costs of 4 h per week were assumed and charged with 35 € h⁻¹. All costs and benefits were calculated for an annual balance with 10,000 t WAS (corresponding to approx. 40,000 population equivalents) and for a WAS with 5% TS and 70% VS.

DOC, COD and K all suggest similar disintegration trends, showing two maxima; a first peak at 70 °C (DOC and DOC) and 85 °C (K), followed by lower values at 120 °C and rising to maximum values at 200 °C (Fig. 1). Looking at the relative values, DOC and DD_COD suggest rather similar degrees of disintegration lying at 26 and 29% resp. for the 70 °C peak and rising up to 54 and 63% for samples treated at 200 °C. For K, the relative disintegration is about 62% at the 80 °C peak and practically the same (63%) at 200 °C.

The reduced disintegration efficiency at 100–120 °C is in accordance with Li and Noike [25], who measured the solubilisation ratio of VS after thermal WAS PT from 30 to 170 °C. Similarly, Kuglarz et al. [26], who measured dissolved COD after thermal PT from 20 to 100 °C, found stable but not increasing COD solubilisation rates as well as decreased ammonia and phosphate release from 70 to 100 °C. One possible explanation for this effect could be Maillard reactions, occurring between amino acids and reducing sugars at elevated temperatures over 80 °C and resulting in polymerization and thus diminished solubilisation of proteins [27].

Although two temperature optima for cell disintegration in the range 15–200 °C are visible, several reasons suggest the application of low temperatures around 70 °C for a thermal pre-treatment; (1) such a PT is energy saving due to the lower supply temperature level. Also, the heat from the combined heat and power unit of the WWTP can be used and the subsequent cooling step can be omitted, as the sludge can directly be mixed with cold primary sludge to a suitable temperature, counteracting the heat losses in the biogas reactor. (2) The PT can be performed at ambient pressure saving investment and operational costs. However, the level of disintegration is not yet fully exploited, leaving space for additional disintegration using an alkaline reagent.

This temperature is also supported by other studies, showing that it requires the lowest energy input to solubilize WAS [26] and stating that it is best in terms of VSS-reduction at a pH of 8 [28]. For all these reasons, a PT temperature of 70 °C was chosen and the following experiments were executed at this temperature.

In this experiment, the thermal PT at 70 °C yielded a DD_COD of 34% compared to the native WAS. This DD_COD was increased by the addition of equal alkalinity as ash (38%), NaOH (46%) and aluminate (47%). While NaOH is known to work well, ash and aluminate have not been tested as alkaline reagents for sludge PT and would have the advantage to be cheap industrial wastes.

When compared to the highest achieved value of the aluminate-treatment, the DD_COD of the thermal and the ash treatments where found to be significantly lower (p < 0.001, −11.6 and −9.2% resp.), while the NaOH-treatment was not significantly different (Fig. 2a). The pH, however, was lowest after the treatment without alkaline reagent (6.82) and increased with the addition of NaOH (7.85), ash (8.45) and aluminate (9.02). We determined NaOH as the optimal alkaline reagent due to the following reasons: if compared to aluminate, it shows the same degradation efficiency but has stable costs that are not subjected to the waste market. Furthermore, the utilisation requisites of waste aluminate are not yet clarified and if the waste product is used, it has a variable composition. Ash is less effective with regard to NaOH, although it has been found to act beneficial to methane concentrations and hydrolytic efficiency in cattle slurry-based biogas plants [29]. However, the here required amounts to reach the desired alkalinity of 0.04 M were rather high and would result in 400 t to be sieved and amended annually. Furthermore, the sedimentation process in the sludge was impacted with increased amendment of ash, lowering the overall economic feasibility of this treatment. Having in mind that additionally, a final pH >8 would probably require a neutralization step of the pretreated sludge [cf. 12] increasing the costs of the pre-treatment, these results suggest the use of NaOH as preferred alkaline reagent.

Compared to the thermally at 70 °C pre-treated sludge (25% DD_COD), the DD_COD increased by 24% using 0.02 M NaOH (49% DD_COD), by 42% using 0.04 M NaOH (67% DD_COD) and by 50% using 0.08 M NaOH (75% DD_COD Fig. 2b), showing a non-linear relationship of DD_COD and NaOH-concentration and suggesting the use of 0.04 M NaOH to be the most efficient in terms of chemical costs vs degradation capacity and being consistent with the optimum concentrations defined in other studies [7, 30].

Applying the optimized PT-conditions 70 °C and 0.04 M NaOH, the dewaterability of the WAS increased from 27.0% TR_A for the native sludge to 32.6% TR_A for the treated sludge, reducing the disposal costs for the remaining WAS substantially. The necessary amounts for the treatment with PE increased from 12.7 kg t TS⁻¹ for the native WAS to 16 kg t TS⁻¹ for the treated WAS.

With the optimized PT-strategy 70 °C and 0.04 M NaOH, the biogas potential test resulted in increased methane production of 213 L_N kg _VS ⁻¹ compared to 174 L_N kg _VS ⁻¹ for the native sludge (Fig. 3), comparable to the production rates observed elsewhere (cf. Table 1). This corresponds to an increase of 22.3% in methane yield. The activity of the microorganisms could be confirmed by their methane production of 362 L_N kg _VS ⁻¹ for microcrystalline cellulose, representing 87% of the theoretical methane potential [23]. These results also confirm that the resulting increased pH of the pre-treated sludge was not inhibiting the microbial consortium. Regarding the physico-chemical parameters, TS, total COD and total proteins where all degraded to a higher extend in the treated substrate (5 vs 13% reduction in TS, 15 vs 18% reduction of total COD and 10 vs 14% reduction in total proteins for native and treated sludge, resp.), while VS showed a similar reduction for treated and native sludge samples (4.2 vs 3.8% reduction for native and treated sludge, resp.).

The cost-benefit calculation shows that the WWTP has overall yearly WAS-related costs of 75,342 € a⁻¹ and that these costs drop by 21% to 59,195 € a⁻¹ with the optimised WAS PT (cf. Table 2). While the costs for heating can totally be retrieved from other WWTP processes by using excessive heat, the most important costs generated by the PT are the chemicals used for the alkaline PT itself (NaOH; 5600 € a⁻¹), the personnel costs (4400 € a⁻¹) and the increased polymer demand for the flocculation of the pretreated WAS (1358 € a⁻¹ increase compared to native sludge). Nevertheless, the increased savings in electrical power resulting from increased biogas production (4714 € a⁻¹ more savings for treated WAS, −22%) and especially the reduced disposal costs (22,791 € a⁻¹ less for treated WAS; −27%) lead to the overall positive effect of the PT (cf. Table 2). With the currently chosen cost parameters, such a PT would lead to total savings of 16,147 € a⁻¹. Not included in the cost-benefit calculation is the investment for the construction of a PT-facility, an investment that would have to be offset by the money saved yearly in less than 15 years, the regular lifetime of such a facility. If estimated with 360,000 €, it will be amortised within 22 years with the here-chosen parameters, causing the PT not to be beneficial enough to be implemented. If, however, the WWTP treats greater amounts of WAS, the PT will turn to be beneficial as soon as these amounts rise above 14,000 t (56,000 PE). Additionally, the overall savings could be enlarged by omitting an often necessary pre-heating of the native sludge prior to the introduction into the digester and by an additional compensation for the thermal transmission losses achieved with the higher temperatures of the pre-treated sludge.