Commissioning of a Modern Centrifuge Solids Dewatering System

Introduction Clean Water Services (CWS) is a public utility providing wastewater, stormwater, and watershed management to the residents of Washington County, Oregon, USA. Among the utility's facilities is the Rock Creek Advanced Water Resource Reclamation Facility (WRRF), a 45 MGD (dry weather) campus that manages the solids stream through mesophilic anaerobic digestion, centrifuge dewatering, and land application of Class B biosolids. CWS has recently completed the design, construction, and commissioning of a modern dewatering system at the Rock Creek AWRRF. Project Overview The existing dewatering equipment consisted of a single Humbolt Bird centrifuge dating to 1990. The size of the machine dictated 24/7 operation to keep up with solids production. Redundancy was provided by keeping a spare rotating assembly (bowl and scroll) on site, and maintenance staff were able to effect a complete change out in the course of an 8-hour workday. The original emulsion polymer make down system was comprised of a Dynablend unit, aging tank with mixer, transfer pumps, a 5,000 gallon feed tank, and 3 dosing pumps. After failure of the Dynablend a single static mixer was used for polymer make down. Over time two additional polymer dosing locations were added upstream of the original location at the centrifuge. Before installation of the Humbolt Bird centrifuge, solids dewatering was done with 4 belt filter presses which were still on site. The project began with removal of the four belt filter presses to make space for the installation of two GEA Watermaster CF 8000 centrifuges. The solids processing building was retrofitted for seismic resiliency as part of a 50-year earthquake preparedness plan. The polymer system was entirely rebuilt, requiring a temporary system using polymer totes. The existing flat bottom bulk polymer tanks were replaced with conical bottom tanks with transfer/recirculation pumps. Three Velodyne VeloBlend VM-P polymer make down units were installed to serve both thickening and dewatering applications and the mixed aging tanks were removed. The ability to use the 5,000 gallon feed tank was retained as an option for both maintenance of the polymer system and providing solution aging. Three polymer dosing points were designed with automatic control and flow measurement where the intermediate dosing point included a static mixer on the solids feed line immediately downstream of the dosing point. The system also included instrumentation to measure the feed solids density (%TS) and the centrate TSS (Valmet, Finland). This instrumentation enables real time cost control calculations and automated polymer dosing to maintain centrate quality. The system is shown in Figure 1. Commissioning The mechanical start-up of the system was impressively trouble free. CWS enjoys the benefit of in-house controls, electrical, instrumentation, maintenance and engineering staff and requires two rounds of operational readiness testing (ORT) that provides early identification and remediation of functional issues. The problem encountered is not at all uncommon and was that the system did not immediately meet performance specification. The performance specification called for the system to achieve 95% capture, and 24% TS cake at a maximum polymer dose of 27 lbs active per ton of dry solids. While the capture could be achieved at the limit of the polymer dose, the cake solids could not be met. While this is not unusual for start-up of a new system, the reduction in cake solids from the old to new system was alarming. The old system was consistently achieving cake solids between 22 and 23%, the new system was fluctuating between 20 and 22% (Figure 2). As with any new system, time is needed for optimization. The operations staff was familiar with dewatering optimization procedures having been trained to tune the process to lowest total cost and began learning the new system. The primary question to answer was in the polymer system. To ensure that the system was utilizing fully activated polymer in the most efficient way possible, a matrix was developed to systematically adjust one polymer variable at a time and analyze performance. Additionally, a static mixer was installed downstream of one VeloBlend to see if there was a difference in more polymer activation energy. The primary variable is the polymer solution concentration and the secondary variables are polymer make up with or without static mixer, injection point, and aging tank. Each condition is run for one day and three sample sets of cake and centrate are analyzed. The matrix becomes complex and time consuming. There are 8 solution concentrations at three dosing points for 24 combinations. Those are tested with and without the static mixer, then with and without aging for a total of 96 conditions. This type of evaluation will not only optimize the process, but find the limit of cake solids at the optimal condition. To date staff have found an optimal solution concentration of 0.24% active over all conditions and insignificant effect of other conditions, except for best results dosing at point 1 which is at the machine. The next test is to rerun the matrix using the aging tank, and adjusting aging times. Summary This abstract, presentation, and manuscript describe a dewatering modernization project incorporating features from the culmination of 10 years of dewatering optimization at CWS. It underscores the pitfalls of performance specifications and guarantees, the time needed for full optimization and elimination of causes of non-performance. By the time of conference the evaluation will be complete and a discussion of resolution of contract, to the degree possible, will be available for inclusion.