What Is the Best Fit

Tuesday, October 1, 2013

Researchers wanted to compare conventional and novel thickening technologies in economic and environmental terms, carrying out the study in three steps. First, they collected operational data from GBT installations from several water resource recovery facilities (WRRFs) in the upper midwestern United States. The process mass balances were assessed by taking the measurements of influent WAS, thickened WAS, and reject water. Data for the process energy consumption and an inventory of the auxiliary process equipment also were collected. Second, the researchers measured the same operational parameters for energy-efficient centrifugal thickening processes at five pilot-test sites in 2012. Finally, economic and environmental assessment of thickening technologies was conducted, based on the data obtained from the field measurements. The researchers applied a life-cycle assessment framework to evaluate the environmental performance of thickening-technology alternatives.

Examining the technologies
GBTs use a porous belt to separate the free liquid phase from the solid phase. GBTs rely on polymers for achieving the required flocculation that frees water from solids, enabling separation. Though little technical data has been published to verify this, GBTs are assumed to use relatively low amounts of energy.

The decanter centrifuge can be used in the thickening process, though it is most common for solids-dewatering applications. Thickening centrifuges apply forces up to 3000G to solids within a rotating bowl horizontally. Thickening centrifuges also are able to perform with reduced quantities of — and sometimes completely without — flocculent polymers, but they are believed to use more power than their counterparts. However, a more-energy-efficient design of a thickening centrifuge appeared on the market. The modification includes the implementation of a near-centerline discharge that has been shown to significantly reduce power consumption and an air-assisted cake-removal system that enables a higher degree of operational control on the total solids content of the thickened WAS cake. According to the 2012 paper, “Thickening Equipment Maintenance Costs” by M. Kopper, this novel design has achieved power use as low as 0.26 kW/m3 per minute (0.06 kW/gal/min), a new benchmark for any type of decanting centrifuge. Kopper also says the novel design also has been shown to work effectively with influent WAS from 0.2% to 2% total solids with and without use of flocculent polymers and demonstrated the ability to control cake solids concentrations in the range of 3.5% to 9%.

GBTs currently tend to be the most commonly installed new application. However, with new thickening centrifuge advancements, the new, energy-efficient design may be an economically and environmentally competitive option for WAS thickening. Little research has been completed in assessing the environmental impact of thickening technologies in the past, either for traditional or novel thickening techniques.

Establishing parameters
Researchers identified several WRRFs for the study and gave each facility a survey to complete, which detailed typical operating conditions. Subsequently, sampling of the influent WAS, effluent filtrate, and discharge thickened WAS was completed. Polymer use was confirmed by measurement, when possible, and power consumption was measured in two parts using an ammeter and voltmeter. Researchers collected power data for the belt drive, the hydraulic steering unit (where applicable), the washwater pump, and motors associated with the GBT room-ventilation system.

The pilot tests were conducted to collect all centrifuge data, with the exception of one site, THK site 1, where the thickenercentrifuge was considered in this study. This was gathered from an on-line installation. In all cases, influent WAS was sampled upstream of the centrifuge, and reject water or centrate and thickened WAS were sampled downstream of the centrifuge, prior to entering storage reservoirs. For each pilot test, samples were collected intermittently by varying bowl speeds and sludge throughputs, with and without polymer addition.

Plant staff analyzed collected samples at each of the respective pilot-test sites, while a consultant recorded operational data, including polymer use and power consumption. The power consumption for the centrifuge consists of three parts: main drive, scroll drive, and air-injection system. It was not feasible to measure the power of the air injection at the time of pilot-testing due to the cycling nature of the air compressor. Because the power for this component is significantly less than for the other two components, values measured at the on-line installation site were applied (normalized based on flow) for the pilot-test sites that used the air system.

Collecting the data
After data had been collected for both the GBTs and thickening centrifuge, the data were compared first in terms of physical performance with the following parameters: increase in the percentage of total solids (TS) of WAS, solids recovery rate, polymer consumption, and power use (see Table 1, p. 58). Percentages of TS of influent WAS and thickened WAS for both technologies were calculated on a percentage-weight/weight basis.

Researchers compared polymer consumption on a pound of polymer per ton of dry solids neat basis because, first, the activity for some of the polymer products was unknown, and secondly, pricing for polymer from vendors typically is calculated on a pound-delivered basis. (It should be noted that all facilities in this study used emulsion flocculent polymers.) Power consumption for each facility was compared on a kilowatt per gallons per minute (gal/min) basis. This was done because influent sludge to thickening equipment typically is relatively dilute (with less than 1% TS), and equipment capacity is thus limited by hydraulic throughput, rather than by solids loading.

Testing showed that with the GBT, the average increase in percent TS of WAS was 4.3%, and 4.0% for the thickening centrifuge. Considering the solids recovery of the two technologies, the GBT sites had a more consistent higher capture rate than the thickening centrifuge, 99.9% versus 94.8%, on average. Despite this difference, the novel thickening equipment offers improved recovery rates over traditional thickening centrifuges. Additionally, when considering the solids capture rate, one also must take into account the flocculent polymer consumed to achieve that level of recovery. On the point, the GBTs used more flocculent polymer than the thickened centrifuge in all cases. One GBT facility in particular, Site B, had an unusually high polymer dose, at 20.8 kg/Mg (41.5 lb/ton) neat. The reason for this high dose is unknown, but it could be attributed to higher quantities of industrial wastewater being treated and possibly to aged equipment. Regardless, the data point was deemed an outlier and excluded from the average and any further calculations. On average, the GBTs used 5.3 kg/Mg (10.6 lb/ton) neat of polymer, whereas the thickening centrifuge on average used 0.5 kg/Mg (1.0 lb/ton) neat. Of the five test sites for thickening centrifuge, two were run completely without polymer. This, it seems, is the new thickening centrifuge’s largest advantage.

Finally, to determine the electrical consumption of the two technologies, power was measured only for system components that were not common to both the GBT and the thickening centrifuge (see Table 2, p. 59). On average, the GBTs consumed about half as much power as the centrifugal thickening technology per gallon per minute (gal/min) of WAS flow, 0.31 kW/m3 per minute (0.07 kW/gal/min) versus 0.57 kW/m3 per minute (0.13 kW/gpm).

Conducting the environmental assessment
The primary trade-offs among thickening technologies are polymer and electricity usage rates. In order to evaluate these trade-offs the research team adopted the life-cycle assessment (LCA) methodology. LCA is a framework that translates environmental emissions to its foreseeable impacts on humans and ecosystems. LCA also considers the embedded emissions for energy and chemicals supplied. In conducting this LCA, global warming potential (GWP) was the only impact category considered, so calculated emissions for all substances were converted to kilograms of carbon dioxide equivalent.

In order to conduct the LCA, researchers used software tool in which users define modules that help determine a complete life-cycle inventory (LCI) of the system. For some of the emissions rates data input, researchers had to import data from the U.S. Department of Energy (DOE) National Renewable Energy Laboratory (NREL).This study primarily considered emissions directly related to the operation of the thickening equipment itself, specifically for process water, chemical addition, and energy consumption. This study did not consider emissions downstreaof the thickening device from the solids stream (TWAS), because it was assumed the device was optimized to control solids and hydraulic loading rates to the anaerobic digester. The study did, however, take into account the energy required to treat the supernatant from the devices, as the GBTs and centrifuges use different amounts of process water, which ultimately adds to the hydraulic loading and, thus, energy use.

Because researchers could not find in any databases emissions data specific to the production of flocculent polymers, they had to use emissions data in the NREL database for major components present in flocculent polymers (ethoxylated alcohols, petroleum distillates from naphtha, and water), with the exception of polyacrylamide. Emission data for the production of a similar polymeric substance, polyarcylnitrile, were used instead. The data were substituted and then adjusted based on the ratio of the atomic masses of the two substances. The assumed consistency was based on the material safety data sheet for a common wastewater flocculent.

The geographical boundary for emissions data was set to the United States. When calculating the emissions associated with the treatment of the supernatant flow, researchers assumed 0.318 kWh of energy per cubic meter of additional hydraulic loading to the wastewater treatment. The rate was based on the findings in Electrical Power Research Institute in 2002. The value assumes a medium-size WRRF, which is approximately 10 mgd (37,900 m3/d), according to the study.

Evaluating the environmental data
Plants using GBTs generally had much higher emissions than the centrifugal-thickening sites (see Figure 1, p. 59). In particular, GBT – Site B had much higher emissions than all other plants for either technology at 68.6 kg of carbon dioxide equivalent per tonne of WAS processed. As previously, the plant was deemed an outlier and omitted from the technology average. Omitting Site B, the average emissions for the GBT sites was 14.1 kg of carbon dioxide equivalent per tonne of WAS processed, while the average for the centrifugal thickening sites was found to be 1.72 kg of carbon dioxide equivalent per tonne of WAS processed.

Using the process view in the life-cycle inventory software, emissions were broken out by module, allowing for better determination of where the most-significant emissions were generated. In both the GBT and thickening centrifuge, the emissions associated with the addition of process water and treatment of supernatant were then found to be an order of magnitude smaller than the combined emissions for use of flocculent polymer and electricity. Because of the way the LCA was constructed, testing which of the technologies have the largest impact was slightly more difficult. Thus, a sensitivity analysis to determine which had the most significant influence on overall process emissions was necessary. To do so, the researchers conducted two separate sensitivity analyses for the average scenario for each technology. The first analysis considered the effect on overall emissions by changing polymer usage in increments of 10%, and the second did the same by changing energy consumption (see Figure 2, p. 60). The researchers then determined definitively that polymer use is the parameter that has the largest effect on GWP for the thickening process for both technologies. It also seems the relationship between polymer use and GWP is relatively linear. The difference in GWP for a particular percentage increase in usage is much larger for GBT sites, because the quantity of polymer used is an order of magnitude higher.

Conducting the economic assessment
The economic assessment was completed in two parts: considering the operational cash flow and considering the 20-year life-cycle costs.

The former consisted of taking the average operational data that was collected and using it to compare the costs of processing 1 Mg of dry solids. The latter uses the average operational data from the centrifugal-thickening pilot tests, but then the team considers each GBT test site independently to complete a cost–benefit analysis for replacement of the installed GBT with the energy-efficient thickening centrifuge. The cost–benefit assessment was conducted using five different techniques: simple payback, net present value (NPV), equivalent uniform annual value (EUAV), benefit–cost ratio (BCR), and internal rate of return (IRR). While simple payback is by far the easiest to calculate, its output does not definitively answer whether the project should proceed. This is why the other four analyses were used. These analyses were carried out based on the guidelines set forth in the 2013 Water Environment Federation’s (Alexandria, Va.) Residuals and Biosolids Workshop, “Using Appropriate Economic Methodologies for Evaluation of Cost-Saving Projects.

Many assumptions were required to complete a thorough economic assessment, including the cost of polymer, the purchasing price of electricity, etc. The cost of polymer varies year to year. Often, municipalities negotiate 1- to 3-year contracts, locking in the price. In the past 2 years, many have seen the cost of polymer decrease. For this study, it was assumed that polymer was purchased by the tote and the cost of that polymer was $0.51 per kg ($1.15 per lb) delivered in 2013; however, an annual escalation factor also was assumed. The factor, 1.19%, was based on an inflation-adjusted increase in the cost of crude oil from 1946 to 2012.

The assumptions. The cost of electricity nationwide varies dramatically in the U.S. However, the national average industrial rate for electricity is $0.066 per kWh, according to DOE’s Information Administration. The national average also is representative of the U.S. Midwest, where the GBT sample sites were located, and, thus, the rate of $0.066 per kWh was adopted as the baseline scenario for 2013. A rate of 2.22% was used as cost-escalation factor for the cost of electricity, also based on DOE Information Administration data. The discount rate was assumed to be 0.8%. While the discount rate used for such analyses varies by location, 0.8% was chosen because it was recommended by the White House Office of Management and Budget for a 20-year projects based on the forecasted value of U.S. Treasury bonds and because it was necessary to compare all sites on an equal playing field. The cost of maintenance for each piece of equipment was assumed to be 5% annually of the equipment’s capital cost. This percentage was based on the testimony of a centrifuge and solids management expert and reflects that centrifuges will have somewhat higher maintenance cost, as the capital-cost-per-machine capacity is higher for centrifuges than for GBTs. In calculating the 20-year life-cycle cost of replacing a GBT with the centrifugal-thickening equipment, a residual value for the centrifuge was assumed. Though the resale value of the equipment varies depending on condition, 10% of the equipment capital cost was conservatively assumed.

Cash-flow analysis results. Using the average values for the five GBT test sites, the average operational costs were calculated on a per dry ton of WAS processed basis. The researchers determined the difference in operational costs between the average scenario for the GBT and the average scenario for the centrifugal thickener (see Figure 3, p. 61). Thought of another way, this difference is equal to potential operational savings that would be had in replacing the GBT with a centrifugal thickener for the average scenario. The graph shows that with the current market conditions (in blue), the potential savings range from $3629 to $9072 per kg ($4 to $10 per dry ton) of solids processed. Additionally, the point where operational costs for the centrifuge exceed operational costs for the GBT was at $0.36 per kg ($0.80 per lb) of polymer, while the price of electricity would have to be $0.20/kWh.

Cost–benefit analysis results. In order to evaluate the feasibility of replacing the existing GBTs with the centrifugal thickening technology now at the five facilities where GBT data were collected, a 20-year project life was used, and the five different cost–benefit analysis methods previously stated were applied. To make the replacement economically feasible, NPV had to be greater than zero, EUAV had to be greater than zero, BCR had to be greater than 1, and IRR had to be greater than discount rate of 0.8%. Simple payback depends on the buyer.

Three sites (A, B, and E) clearly meet the criteria for project implementation, while sites C and D fell short of paying for themselves within the 20-year project life considered (see Table 3, above). Knowing that polymer cost had the highest influence on possible savings by replacement with the energy-efficient thickening centrifuge, there must be a threshold quantity of polymer consumed annually at a facility using a GBT for which a project would break even at 20 years. To determine this quantity, the researchers used the Solver data analysis tool in Microsoft Excel to determine the amount of annual polymer use needed to set the NPV to zero for the average scenario. This value was 19,800 lb (9000 kg) of polymer annually per GBT. Considering the amount of solids processed for the three feasible sites, this would be 6.4 lb (2.9 kb) per 5.3 Mg (1 ton neat) as a minimum dose for a GBT facility. Following this exercise, a similar analysis was completed to determine the minimum amount of solids processed annually per GBT to merit replacement with the centrifugal-thickening device. Again using the Solver and assuming the average polymer dose found for the GBT sites excluding Site B) of 23.3 kg (10.6 lb) per 5.3 Mg (1 ton neat), the minimum quantity of solids processed annually to merit replacement was 2277 dry Mg (2070 dry ton) per year per GBT.

Study limitations
While completing the study, the researchers encountered a number of limitations, meaning there were areas where work could be improved. First, the number of sites referenced is small — only five for each technology. A sample size of 20 to 30 plants would be far more desirable; however, such work will take considerable time.

When considering the data collected, a mass balance was taken across the process for each facility. Though increases in percent TS were given, a more valuable metric would have been percentage of time that discharge TWAS was maintained within the desirable range. The reason for such an analysis would be because many facilities have a target percentage TS of TWAS, and maintaining consistent operation near that target is considered important for thickening equipment, as devices frequently operate 24 hours a day, and slight fluctuations can have major impacts downstream, such as for digester heating requirements. Unfortunately, recording such data would have been substantially more difficult.

For the environmental assessment, the biggest limitation was the LCI data available currently for flocculent polymers. While it is unlikely that embedded emissions associated with electricity consumption would outweigh the impacts associated with the production of polymer, the uncertainty for the magnitude of embedded emissions for polymer production is quite high. Thus, future research should work to better account for the production of the flocculent polymers.

Finally, considering the economic assessment, the primary limitation was the assumptions used for maintenance costs of equipment. Annually, 5% of capital cost was assumed for the maintenance of each technology. Because the new centrifuge has only been in the market for a short time, long-term maintenance can only be estimated based on traditional centrifuge designs. However, this equipment, in theory, should require lower maintenance over its lifetime, because the equipment typically runs at 50% to 70% of the speed of conventional centrifuge designs.