Study draws to a conclusion
Results of an extensive study of nine WPL SAF plants are revealed
The study criteria focuses on six main areas of interest:
- wastewater characteristics and flows,
- operational problems,
- control and maintenance,
- operating costs and power consumption,
- recovery from shutdown,
Tighter environmental regulations make it necessary for secondary treatment of domestic wastewater at plants dealing with a population equivalent (PE) of less than 2,000. As part of this programme Yorkshire Water Services has monitored the performance of nine WPL SAFs with sizes between 20-250PE.
Small treatment systems are not a scale-down of larger systems. In contrast to their larger counterparts, issues of short-circuiting, solids breakthrough and prolonged periods of no flow with intermittent high peak flows have potential to reduce performance and need to be addressed.
A package sewage treatment system should provide advanced treatment to meet a range of different standards, accommodate large variations in hydraulic and organic loadings, be simple and easy to operate, be cost effective and be regularly operated and controlled by trained personnel.
Sites B, D, G and I were monitored intensively, with composite samples taken of the raw sewage, settled sewage and effluent on a daily basis. Spot samples were also taken from every other biozone compartment and flow levels were recorded. Weekly samples were taken from the remaining five sites.
The four sites visited frequently were monitored for flow and BOD throughout the day which gave a diurnal flow and load profile. The results indicated a typical peak in flow during the morning with a lower peak in the evening. The only exception was Site G which had a pub in the catchment area and here the evening peak was higher. See Figure 1. The strength of BOD and ammonia in the wastewater also showed similar peaks.
The smaller the plant, the more accentuated the peaks, in terms of maximum peak to average flow ratios. However, despite the high peaks at Site B, performance was reliable.
The wastewater analysed was typical medium strength domestic sewage (as defined by Metcalf & Eddy), ie BOD at 185mg/l and ammonia at 39mg/l. In all cases there was adequate pH and alkalinity to allow nitrification to proceed.
The wastewater characteristics showed sharp diurnal variations with concurrent flow and strength, morning and evening peaks. There was no sign of a third early afternoon peak as has sometimes been suggested.
The scale of variation in the flow of pollutants varied inversely with the size of the community. Overall the SAFs coped well with the variations in loading, as reflected in minimal diurnal final effluent variations. In all cases estimated flows and loads were adequate to ensure sufficient plant design. Accuracy in estimated flow data increased with larger plants. There was no indication that oversized plants caused performance problems.
Some operational problems which affected performance were identified.
The study also surveyed operators and managers. The sites were reported as "lovely" and "easy to operate". Suggestions for improvement included automatic restart. The tanker drivers were equally happy and reported a de-sludging time of approximately 15 minutes. Little information could be found with regard to real maintenance costs, partly due to lack of recording and little requirement for maintenance work.
Small treatment plant failure is generally due to poor design for the conditions and mechanical wear and tear. The SAF survey did not discover any failure modes directly related to poor flow estimations and unsuitable design. Any problems were linked to lack of operational control.
The visit regime for sites is less frequent than recommended for small-activated sludge, trickling filters and rotating biological contactors (RBCs). The extended interval can be justified on the basis that a SAF is considered more biologically robust than the activated sludge and trickling filter processes, and more mechanically reliable than the RBC process.
WPL's design is generally accessible, with large lids enabling easy visual inspection. However, an efficient operator will only spend 15 minutes on-site per inspection, so it is felt the airlifts should be set with no more than a 10-minute off-time, so it can be checked during a routine visit.
Figure 2 displays the power consumption per PE on a daily basis and indicates that economies of scale take place even between plants ranging from 20-250 PE.
There is a clear per capita saving on power consumption as the plants increase in size. It is evident that SAFs use more energy than RBCs. However, Fastenau et al (1990) reported a breakdown frequency for RBCs in the range of one-three times/pa. The savings in power consumption are, consequently, offset by the cost of shaft breakdowns and replacements.
Figures 3 and 4 indicate that on average the plants operated within the descriptive consent.. Most plants were capable of achieving it under normal operating circumstances. Figure 5 shows the average performance of the intensively studied plants in terms of suspended solids removal. The figures indicate the plants perform reliably when their operation is not interrupted, for example, by a power cut.
The plants were not designed to nitrify but in most cases there was sufficient capacity, due to design safety margins, which allowed some nitrification to take place. In the case of the three plants with no performance interference, there was sufficient capacity to nitrify to 10mg/l ammonia with a 0% chance of a 95%ile failure, demonstrating the ability to nitrify was reliable.
Power cuts caused odorous sub-standard final effluents within 24h. Figure 6 demonstrates how the biology at Site G recovered after a five-day period without aeration. It is clear the less sensitive carbonaceous oxidation process meets the consent within 24h. Similarly, suspended solids decreased to below 30mg/l within 48h. The ammoniacal nitrogen (ammN) removal had a much longer recovery time. At Site G, a 14-day period was necessary for the ammN levels to fall below 10mg/l. At Site F the recovery period for ammN was nine days after a shut-down period of eight days, and at Site E the recovery period was two days after a shut-down period of three days. All plants mentioned had capacity to nitrify to varying degrees.
In all cases plants performed within consent when no operational incidents interfered. Biological treatment was the most reliable with the plants more susceptible to solids breakthrough than BOD or ammonia. Despite the high peak flows experienced by smaller plants, performance was maintained with low-zero probability of breakthrough in BOD and ammonia.
Primary tanks performed as expected with 30-40% BOD reduction. Plants with separate sludge storage had reduced efficiency, possibly due to the return of particle rich liquors to the settlement tank.
Where existing septic tank primary tanks had been used, BOD reduction was comparable to the WPL primary tanks, but de-sludge took longer. However, ammonia levels tended to be higher from the septic tanks.
Biozones held dissolved oxygen (DO) levels above 2mg/l and pH was maintained
between six-nine. DO levels rose across the segmented biozone as BOD, ammonia
and alkalinity levels dropped. The final settlement tank removed between 60-99%
of suspended solids, although on some occasions soft flocs of biomass were noted
in the final effluent under normal operating conditions. Overall the SAFs accommodated
load variations well, as reflected in minimal diurnal final effluent variations.
The report identified the following advantages and disadvantages associated with SAF technology:
- no recirculation of flow or pumping of backwash water,
- no moving mechanical parts,
- little surplus sludge production due to high micro-organism bio-diversity and long sludge retention time,
- the process can sustain and adapt to fluctuations of hydraulic and organic loading since it possesses a larger amount of biomass and a longer food chain compared to activated sludge,
- the long retention of biomass solids gives good protection against toxic substances,
- high surface-to-volume ratio of support media offers the possibility of a significant reduction in volume, compared to suspended-growth systems.
- the process is not maintenance free,
- small mechanical units are more susceptible to breakdown than large units,
- high turbidity in final effluent can occur because of tiny particles of broken biomass that does not settle well,
- medium-term extra cost,
- poor sludge stabilisation compared to suspended growth processes,
- slow start-up and time needed before full recovery after system breakdown,
- little flexibility once designed,
- some odour problems have been reported,
- no correspondence in actual air requirement and air supply throughout different stages of the process,
- accumulation of fats, oils and greases can cause clogging problems,
- detergents can lead to foaming problems,
- suspended growth systems have better oxygen transfer efficiency,
- tank dimensions are limited by the need for transportation,
- compactness can lead to more turbidity and higher temperature effects.
The report concluded:
- in Yorkshire Water Services the estimated wastewater characteristics safety
ensured the plants were reliably
compliant to their discharge consent,
- existing primary septic tanks could be used for BOD removal systems,
- separate sludge storage tanks reduce the efficiency of the primary tank,
- none of the evaluated plants were overloaded,
- overall consent compliance tended to be compromised by operational or M&E problems, rather than improper process design,
- the SAF plants evaluated had minimal maintenance costs,
- new installations should be sited as far from local residents as possible,
- frequency of maintenance, control and mechanical breakdown will affect the overall success of the SAF plants,
- improvement recommendations to the SAF include auto-restart in working order and feed forward airlift should be operated at intervals of less than 15 minutes,
- a SAF is considered more biologically robust than the activated sludge and trickling filter processes, and more mechanically reliable than the rotating biologicalcontactor process.
Commenting on the study, Richard Munden, managing director of WPL, said:
"Being involved in the trial has been a very worthwhile experience. We're delighted with the results. Customers regularly advocate the design and reliability of our plant and the trial has reinforced this message. With regard to operational events and their implications on performance, this is an area, as a manufacturer, over which we have no control. However, I am sure that the comments relating to this issue will be extremely valuable to our customers.
We also read with interest the improvement recommendations and will seriously consider them as part of our ongoing research and development programme. Customer feedback is always welcome, as we strive to constantly improve our product range."