Henriksdal wastewater treatment plant (WWTP) in Stockholm is currently being extended and rebuilt for increased capacity (from 0.8 to 1.6 million PE) and enhanced treatment efficiency (6 mg TN/L, 0.20 mg TP/L, 5 mg BOD7/L).
The reconstruction includes retrofitting of the existing conventional activated sludge (CAS) tanks with a new membrane bioreactor (MBR) process containing 1.6 million m2 of membrane area. It also includes extended pretreatment and a new treatment step for thickening of primary sludge. Digestion of thick sludge (~6 % TS) will be done at thermophilic conditions, unlike today’s mesophilic operation, with high organic load and relatively short retention time.
To increase the knowledge of MBRs in Nordic conditions, Stockholm Vatten och Avfall (SVOA) and IVL Swedish Environmental Research Institute have conducted long-term MBR studies in pilot scale at the R&D-facility Hammarby Sjöstadsverk, located on the premises of the Henriksdal WWTP. The MBR-pilot was taken into operation in 2013 and was reconstructed to its current configuration in 2016. In 2017 the MBR pilot was supplemented with a sludge treatment line to study different aspects of sludge digestion.
During 2021, the MBR-pilot was operated at a fixed inflow of 4.1 m3/h, which is 37 % higher than the design average flow, with externally provided glycerol as well as internally produced VFA as carbon source for post-denitrification. Aluminum (PAX) was used instead of Ferric (PIX) as complement to Ferrous (FeSO4) for phosphorous precipitation. This was done to test the operational strategy for the first MBR line in Henriksdal WWTP. The average effluent concentration of nitrogen and phosphorus was 3.9 mg TN/L and 0.07 mg TP/L, respectively, which means that the effluent requirements were met also this year. To achieve this, 8.6 g Fe2+/m3 and 0.9 g Al3+/m3 was required.
During flux enhancer trials a total of 17.8 g iron (Fe2+ + Fe3+)/m3 was added. The glycerol dose was equivalent to 17.3 g COD/m3 and for internally produced VFA the dose equivalent was 15.5 g COD/m3. The slightly higher consumption of phosphorus precipitation chemicals compared to 2020, 1.29 mole metal per mole of phosphorus removed, was mainly due to a lower enhanced biological phospho¬rus removal (EBPR) activity in 2021. In 2021 the phosphorous release rates were low during the spring and showed < 1 g PO4-P/kg VSS,h in June but recovered in the summer with 5.5 g PO4-P/kg VSS,h in July after the defoaming agent dosing was stopped.
The iron and aluminum content in the activated sludge was 6.2 and 0.7 %, respectively. Average total sludge age during 2021 was 17.2 days and average aerated sludge age was 7 days. Nitrification was always complete with ammonia concentrations below 2 mg/L except week 25. Test with use of internally produced VFA as carbon source showed that the specific COD consumption was almost the same as for glycerol when comparing the yearly average from 2021 and 2020. Effluent nitrate and total nitrogen removal was similar during the trial with VFA as the rest of the year, when glycerol was used.
Like previous years, the membranes in membrane tank 1 (MT1) was cleaned with oxalic acid and the membranes in MT2 with citric acid. Both membranes were also cleaned with sodium hypochlorite. The membranes were operated with an average net flux around 21 to 25 L/(m2·h) but starting from week 25, the flux was increased to 30 L/(m2·h) which is the design net max flux of the full scale MBR in Henriksdal and was tested in the pilot for 25 weeks.
The net TMP varied between 49 and 218 mbar for MT1 and between 51 and 146 mbar for MT2. TMP was reduced after each recovery cleaning (RC) with hypochlorite, but the effect did not last long. The permeability was generally above 200 L/(m2·h·bar) throughout 2021-2022 for both membranes. Recovery cleanings were done twice with hypochlorite and once with acids during 2021. During 2022 a final RC, first with hypochlorite then with acids was carried out.
The first RC for MT1 resulted in a clear increase in permeability after cleaning. For MT2 the major increase in permeability was the result of a citric acid MC (one week after the hypochlorite RC). The RCs at the end of 2021 and in March 2022 had clear but smaller positive impact on permeability. Prior to the first RCs, permeability was higher for MT1 (cleaned with oxalic acid) compared to MT2 (cleaned with citric acid). After the first RCs, both membranes had similar permeability. As a result of the tough operational strategy from week 25 2021, permeability decreased quite quickly after RCs. MT2 reached a stable level around 250-300 L/(m2·h·bar) while MT1 decreased even more, to as low as around 200 L/(m2·h·bar).
Emission of chlorinated compounds in the off-gas ventilation were measured during the final sodium hypochlorite recovery cleaning. The emission process was slower than expected and generally no clear sign of attenuation of emissions was observed during the 21 hours of sampling. Although composite samples of several hours during the night are not providing enough details, it was concluded that the emissions can be harmful during the entire RC process from an exposure perspective. Trichloramine peaked at 36 times the recommended limit, chlorine gas at 73 % of the short-term exposure limit (15 min exposure), and chloroform at 9 % of the occupational exposure limit (8-hour workday average).To follow up previous measurements of greenhouse gases nitrous oxide (N2O) and methane (CH4), a new campaign was performed during several months in 2021. Generally, emissions observed in 2021 were significantly higher than in previous campaigns in the pilot and especially high N2O-emissions from the membrane-tank could be identified.
No clear reason could be identified but the increased incoming load with a maintained effluent quality and a “better” sampling setup may partly be an explanation.In collaboration with Kemira, tests with a flux enhancer product were performed in 2021. However, no obvious positive or negative change in permeability due to dosing of flux enhancer was possible to identify based on continuously monitored process parameters and commonly observed variations in permeability and effect of membrane cleaning. As the formation of foam is a common phenomenon in MBR plants, tests with an antifoaming agent were done by dosing in batches and continuously to the biological treatment during the period of heavy foaming (March-June).
Even if foaming was not avoided, a good reduction and control of foaming could be achieved. An optimal effect was achieved with continuous dosages of > 10 ppm. However, even though the product has shown to have a positive effect in the MBR-pilot, a permanent use in full-scale may not be economically feasible due to the high consumption.Test with a reduced RAS flow from the design value of 4×Qin to 2×Qin was done with the aim to reduce energy consumption for pumping. A reduced RAS flow would however imply an increased sludge concentration in the membrane tanks, which may have negative effects on the membrane performance with more clogging and consequently increased aeration for membrane scouring and need for more frequent membrane cleaning.
However, no negative effects of the reduced RAS-flow could be seen on the membrane performance. During 2021, tests with a transition from mesophilic to thermophilic anaerobic digestion, dewatering of digested sludge after mesophilic and thermophilic digestion, and thermophilic digestion at high organic loading rate (OLR) and low hydraulic retention time (HRT) were performed in the sludge pilot. Results show that the transition from mesophilic to thermophilic digestion can be done without any major problems if the load was reduced during the most critical temperatures and that stable operation was achieved after 10-12 days. Evaluating the dewatering of mesophilically and thermophilically digested sludge was more difficult and no clear differences could be observed. However, it was concluded that used methods for determining dewaterability or optimal polymer dose are not reliable.
Trials with high organic loading rate at thermophilic digestion showed that the digester performance could be maintained up to an OLR of around 4 kg VS/m3, d and an HRT of 12 d. When the load is further increased and HRT decreased, the performance in terms of VS reduction and biogas-/methane production decreased although the reactor operation was still stable. The overall resource consumption in the pilot showed that the consumption of glycerol was the same as for the future Henriksdal design, even though the nitrogen load in the pilot was 21 % higher and the average effluent total nitrogen concentration was 3.9 mg TN/L compared to the design of 6 mg TN/L.
Also, the iron/metal consumption was 73 % of the future Henriksdal design, although the phosphorus load to the pilot was about 50 % higher compared to design values and effluent phosphate concentrations were below the target concentration. This is mainly explained by the EBPR activity in the pilot. Also, the consumption of cleaning chemicals was lower than the future Henriksdal design although the inflow to the pilot was 30 % higher than design.