OPTIMISING SEWAGE PUMPING STATION PERFORMANCE

Transcription

1 OPTIMISING SEWAGE PUMPING STATION PERFORMANCE Andrzej Krawiec, Martin Byrnes Sydney Water Corporation, Sydney, NSW, Australia ABSTRACT At Sewage Pump Stations (SPS) grease, fat and other material such as wet wipes combine on the well surface over time and can form a significant crust, which causes many operational and maintenance problems. Level sensors and instruments become blocked and malfunction, pumps are choked by this material and odour builds up in the material. Traditionally maintenance crews go to SPS on a regular basis to manually clean the wet well using hoses. Issues with this manual cleaning process, or flushing, include high labour costs, safety risks and confined space entry, isolation of the SPS, pump chokes, ineffective cleaning and excessive potable water usage. This time consuming and labour intensive process is very expensive with each SPS requiring on average 5 to 6 manual flushes per year. Sydney Water has devised and is implementing an automated process of Auto Flushing or selfcleaning of wet wells on all Submersible SPS. This process will largely eliminate the need for manual flushing, significantly reduce odours and optimise performance of the Sewage Pumping Station. To quantify the potential benefits of the Auto Flush process to Sydney Water a 12 month trial at 17 Sewage Pumping Stations was undertaken. The outcome from this trial was so successful that Sydney Water has decided to implement the Auto Flush process across all 500 Submersible Sewage Pumping Stations. When fully implemented this program is estimated to reduce Sydney Water s operating costs by at least $1M per annum. This paper focuses on the results of this trial and will detail how Sydney Water devised and implemented this solution in a simple cost effective manner requiring no capital outlay on new equipment or site construction works. METHODOLOGY The crusting problem faced by sewage pump station operators is illustrated in Figures 1 and 2. The well surface crust consists of non-dissolvable material like wet wipes combining with fat to form a solid mass. Crusting combined with silt build up on the well floor means there has been a regular requirement to manually flush and occasionally dredge sewer pump station wells. This expensive, time consuming and possibly dangerous task forms a significant portion of the on-going operating costs of any sewage pump station. Sydney Water operations came up with the idea that if you started with a clean well you could keep it clean by running the pumps down to a low level near the top of the pump impeller so that the pump would skim off the surface material. If you did this on a daily basis then there would be no extra burden on the pumps and there would be no build-up of material and the well would remain clean. Running the pumps at a low level in the well, combined with the inflow and well shape, also creates additional turbulence which causes the floor silt to be swept up and drawn through the pumps. This in turn leads to no build-up of material on the measuring instruments and large lumps no longer form to choke the pumps. Without build-up of surface material and anaerobic conditions odour is significantly reduced.

2 Figure 1: Before Auto Flush a heavy surface crust develops within 3 weeks of manual flushing (SP0085) Figure 2: Before Auto Flush at one of the earliest trial sites SP0306 showing heavy crusting on the surface.

3 The key to making this happen was to get a reliable and effective process incorporated into the automatic control system Remote Terminal Units (RTUs) that control each SPS. In control terms the Auto Flush operating regime lowers the pump cut-out from its normal level high on the pump body down to the top of the pump impeller, which means the flared intake of the pump will gather in surface material and the pumps will actually stop around the snore level of the pump. Figures 3 and 4 illustrate this revised operating process. The suction trip protection for the pump is typically set at the bottom of the pump impeller so it is necessary to lower this a few percent in the well to avoid nuisance suction trips. In addition a delay time of 20 seconds was typically added to the suction trip to allow for the pump slow down period. The RTUs are part of the Sydney Water Integrated Instrumentation Control, Automation and Telemetry System (IICATS) which monitors and controls all the water and sewer network assets of Sydney Water. One of the key features of IICATS is that it allows automatic time scheduled adjustment of almost any operating parameter. Within IICATS all the pump control cut-in, cut-out, alarm and suction trip levels are derived from set points (limits) on the analogue control level. Using the IICATS ability to automatically time schedule changes to operating parameters such as these analogue set points (limits) the cut-out levels of the pumps are lowered to the same level as the normal suction safety level for a number of pump cycles in the peak inflow period for each SPS. All pumps at the SPS should run down to this low flush level each day to gain maximum effect from the Auto Flush process because each pump will draw in surface material and silt from its immediate surround, so in order to effectively clean the whole well daily, each pump needs to get a run at low level to clean its area of the well. Typically this will mean that low level operation will extend for 1 or 2 hours daily, depending on the number of pumps, site inflow and speed of well cycles. Efficiency of cleaning the whole well is impacted by structures within the well, position of pumps and angle of inflow, the less clutter in the well the better the cleaning. For some particularly problematic sites this Auto Flush period is repeated in the evening peak, or more often if necessary to keep the well clean. Although not part of the initial trial, the Auto Flush process has subsequently been successfully implemented on variable speed pump stations as well as the standard fixed speed pump stations. Variable speed pumps control to a level set point to adjust speed to match the site inflow, this control level set point was also lowered to just above the pump intake under time schedule control to ensure the pumps operated at low level even when the cut-out was not reached. The Auto Flush process was moved to low flow night periods on those large variable SPS where peak day time flows were too high to achieve effective control at low levels in the well.

4 Submersible Pump Auto Flush Cut-out Figure 3: Revised operating levels for Auto Flush. Figure 4: Typical IICATS screen with trend of wet well operation showing Auto Flush. TRIAL SETUP A trial for the Auto Flush process was commenced in February 2014 consisting of 17 submersible SPS spread across the Sydney Water area of operations with particular focus on choosing SPS that had persistent problems of fat build up, crusting and silting for the trial. Some of these trial sites had experienced persistent odour complaints and were ideal candidates to determine if the Auto Flush process would be effective. Setup of the trial sites at each SPS consisted of the following steps: 1) Do a manual flush of the well to remove existing solid crust. 2) Confirm the correct operating levels for Auto flush on site by direct observation. 3) Implement the time schedule control in the IICATS system and RTU. 4) Test operation of Auto flush and adjust levels or suction trip delay if necessary to avoid pump trips.

5 5) Carry out any re-calibrations to level sensors or pump repairs highlighted by the trial. RESULTS Once the trial had been in place at least a month each SPS was visited to confirm the effect. The following observations were made: 1) Floating wet wipes and fat had not accumulated at trial sites. 2) There had been no fouling or faults with instrumentation. 3) Pumps had not experienced chokes. 4) There had been no odour complaints. 5) Pump reliability had improved as measured by fewer breakdowns. 6) Reduced occurrence of suction trip alarms. Figures 5 and 6 illustrate the improved state of wells after Auto Flush has been in place for some time. Figure 5: After Auto Flush the surface remains liquid and free of solid crust 4 weeks after implementation (SP0085)

6 Figure 6: After Auto Flush at SP0306 shown in Figure 2 before shot now showing a liquid well surface. The most immediate benefit of an Auto Flush process is the reduction in labour costs associated with manual well flushing and no flow isolation or flow management which may be required when doing manual flushing or dredging. This reduction is clearly displayed in Figure 7 which is the quarterly costs of preventative maintenance jobs at all 17 trial SPS combined. Before the Auto Flush trial this cost was averaging around $20,000 per quarter for the 17 SPS, due in large part to manual flushing and dredging costs. Now post implementation the cost is trending around $5,000 or less, a 75% reduction. These figures were obtained from the Sydney Water maintenance database and an analysis of the database shows that the Sewer Pumping Stations are flushed on average five times a year at an average labour cost of $400 per flush. It has been estimated around 500 stations in Sydney Water can be operating with an Auto Flush arrangement, saving $1M per annum in labour costs annually for manual flushing alone, not including the other additional benefits of reduced breakdowns and equipment failure. In sewer pumping stations where weekly flushing is conducted, which does happen on some large variable speed SPS, the initial cost of implementing Auto Flush can be recovered during the first month of operation. Other benefits include the reduction in frequency of pump chokes which would have been caused by the gradual accumulation of solids in the well. Each pump choke can cost thousands of dollars to repair. This reduction in breakdown costs is displayed in Figure 8 which is the quarterly costs of all breakdown jobs at all 17 trial sites combined. This cost was averaging around $45,000 per quarter for the 17 SPS before the Auto Flush trial and now post implementation the cost is trending around $15,000 or less, a 66% reduction. The breakdown costs cover all jobs including instrument failure which has reduced also since the introduction of Auto Flush because the

7 instrumentation is not being fouled by the build-up of wet wipes and fat in the well. Other benefits that are not measured financially include the potable water saved by not doing manual flushing. It is estimated that the amount of water used to flush the 17 trial sites is roughly 600 k/l per year. Similarly, the reduction in odour complaints has been noticeable at the 17 trial sites because odour is no longer created by old built up material that forms the well crust. This in turn may lead to a further reduction in cost if less chemical dosing is required to counteract odour in the system. The net result of the Auto Flush self-cleaning process is that the total operating costs for the SPS fall dramatically to around 25% of what they were before the trial. This result is displayed in Figure 9 which is the quarterly total operating costs (preventative + breakdown) for all 17 trial SPS combined. This was trending around $95,000 before the trial and is now below $20,000 post implementation. This result has been achieved consistently across the trial sites and the reduction in maintenance and operating costs is still improving as the cumulative effect of a selfcleaning sewage system flows through and also improves the performance of Sewage Treatment Plants at the end of the network. The final benefit that should not be forgotten is the improvement in safety that is derived from not having to access the SPS during manual flushing and dredging operations. Personnel are no longer at risk of accident or injury in the confined and difficult working environment of a sewer wet well when carrying out these operations. Figure 7 Preventative Maintenance Cost trend for 17 trial sites. Includes manual flushing and dredging costs.

8 Figure 8 Breakdown Cost trend for 17 trial sites. Includes breakdown jobs for all site equipment. Figure 9.Total Cost trend for 17 trial sites. Combines Preventative, General, Corrective and Breakdown Cost. CONCLUSION The main benefits of the Auto Flushing process are: Reduces significantly or completely removes the need for manual cleaning. Reduces breakdown jobs because of less chokes and instrument failures. Improved safety as manual cleaning and confined space entry are reduced. Reduced well silting and less dredging required. Reduced odour with possible reduced need for chemical dosing. Reduces operational issues and cleaning costs at Sewage Treatment Plants. The results indicate that regular removal of floating material on a daily basis with this Auto Flush behaviour is reducing SPS operating costs by 75% and markedly improving reliability of the SPS and the sewer system as a whole. This process provides a low cost, self-sustaining, environmentally friendly solution for keeping the well clean while improving aerobic activity, all without the use of expensive spray systems or potable water.