Water Journal : Water Journal March 2011
sewer processes refereed paper technical features 86 MARCH 2011 water Discussion From the trends illustrated in Figure 8, it is evident that the optimised Caustic Wash was able to maintain dramatically reduced H2S levels for a period well in excess of 20 days (at which point the LAP had to be shut down for other operational reasons). Unfortunately, since the LAP is critical operational infrastructure, we did not have the luxury to control its operation purely for the benefit of obtaining an ideal set of experimental results. Hence, the precise duration of these reduced H2S levels, under controlled conditions, is undetermined. The trend from Wash #3 (Figure 8) suggests that several months of good results may be achieved from an optimised Caustic Wash. It was found that after several washes, both the magnitude and duration of the H2S decrease significantly improved. After Wash#3, H2S levels remained quite low for a period in excess of in excess of 180 days, which is in support of Figure 8. At 180 days (six months), the H2S levels were still approximately 60% lower than the pre-wash levels. This is a significant achievement, even given the fact that the cooler temperatures and other network improvements may have partially aided in this reduction. In addition to these results, long-term H2S monitoring in the downstream gravity sewer main (which receives the effluent and sewage flows from the LAP) has indicated a notable reduction in H2S levels after the Caustic Wash, as was expected, which is a welcome added benefit. While these results indicate a significant reduction in H2S gas, the measurement and statistical analysis of methane levels has been more difficult. Methane, a combustible gas, is measured in terms of 'lower explosive limit' (LEL), and these sensors had been somewhat problematic, requiring frequent calibration. Even with the biofilm-enhanced methane production, the pre-Caustic Wash levels were generally quite low, with the occasional high LEL spikes being the main concern. Because of this, and the frequent calibration problems, the direct analysis of methane data was not possible. However, since it is a similar biological process that produces both H2S and CH4 we can infer that the Caustic Washing and biofilm removal has had a similar effect in significantly reducing the levels of both gases. As well as this, methanogenic bacteria are far more sensitive to high pH conditions than the sulfate-reducing bacteria (Gerardi, 2006), so the observed reduction in H2S levels indicates an even greater reduction in methane levels. In support of this, the frequency and magnitude of high methane spikes was observed to decrease after the Caustic Wash, thereby improving operational safety of the network. During the research and planning period for this Caustic Wash, Sydney Water assessed several other options for the LAP. Other possible solutions included further construction of air treatment facilities in key locations, or operating online chemical dosing of ferrous chloride, calcium nitrate or magnesium hydroxide. The capital cost and operational costs of magnesium hydroxide was reviewed, as it was the most favourable option; however, this was likely to exceed one million dollars and does not offer any exceptional guarantee of odour control, especially at points of turbulence. By comparison, the capital cost of caustic washing was around $40,000 and the cost per wash was around $4,000. Based on the current dosing rig design, now that the initial work is done, this rig could now be constructed again for a significantly lower capital cost, or simply unbolted and relocated to any desired location. In addition to the direct H2S and CH4 monitoring, measurements of solids at the inlet of the receiving Malabar WWTP, and of pumping pressure at the Liverpool pumps, appear to support the removal of biofilm. A spike in volatile solids coincided with the alkaline flow as it entered the WWTP, jumping from 400mg/L to 1000mg/L; however, this did not harm the WWTP treatment processes as it was only for a brief period of time. Pressure at the Liverpool pumps also dropped by up to five metres head after the washes. This finding merits Figure 8: Magnitude and duration of effect for higher dose pH Caustic Washes.
Water Journal April 2011