Implications of Transitioning from De Facto to ... - ACS Publications

Apr 14, 2016 - United States, thermoelectric power plants account for 41% of the freshwater ... however, the additional heat loading increases natural...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Implications of Transitioning from De Facto to Engineered Water Reuse for Power Plant Cooling Zachary A. Barker and Ashlynn S. Stillwell* Civil and Environmental Engineering, University of Illinois at Urbana−Champaign, Urbana, Illinois, United States S Supporting Information *

ABSTRACT: Thermoelectric power plants demand large quantities of cooling water, and can use alternative sources like treated wastewater (reclaimed water); however, such alternatives generate many uncertainties. De facto water reuse, or the incidental presence of wastewater effluent in a water source, is common at power plants, representing baseline conditions. In many cases, power plants would retrofit openloop systems to cooling towers to use reclaimed water. To evaluate the feasibility of reclaimed water use, we compared hydrologic and economic conditions at power plants under three scenarios: quantified de facto reuse, de facto reuse with cooling tower retrofits, and modeled engineered reuse conditions. We created a genetic algorithm to estimate costs and model optimal conditions. To assess power plant performance, we evaluated reliability metrics for thermal variances and generation capacity loss as a function of water temperature. Applying our analysis to the greater Chicago area, we observed high de facto reuse for some power plants and substantial costs for retrofitting to use reclaimed water. Conversely, the gains in reliability and performance through engineered reuse with cooling towers outweighed the energy investment in reclaimed water pumping. Our analysis yields quantitative results of reclaimed water feasibility and can inform sustainable management of water and energy.



INTRODUCTION AND BACKGROUND Reliable energy and clean water are tantamount to a high standard of living in the modern age. As we try to meet the growing demands for water and energy, we do so under increasing environmental and political stress.1,2 Researchers continue to study energy and water resources independently as well as jointly to mitigate stress under these conditions.3 Previous work demonstrated that the energy and water sectors positively and negatively interact with each other, a connection commonly known as the energy-water nexus.3−16 In some cases, such as conservation or resource recovery, the two sectors can be synergistic.4,7,12,17 Conversely, trade-offs can exist where efficiency in one area might increase consumption in the other, such as increased energy consumption for distributing reclaimed water (wastewater treatment plant effluent) through a network.18 We focus our analysis of the energy-water nexus on thermoelectric power plant cooling. Previous work on this topic primarily focuses on quantifying the existing and future power plant water demand in a changing environment,14,19−29 with increasing focus on alternative water resources. In the United States, thermoelectric power plants account for 41% of the freshwater withdrawals and 3% of the consumption.30,31 As a result of these studies, renewable energy sources, such as wind and photovoltaics, which require no water for operation, are attractive from a water perspective. © XXXX American Chemical Society

Increased scrutiny on power plants has led to policy developments such as the Existing Facilities Rule and Clean Water Rule that build on the Clean Water Act, and the Clean Power Plan that sets performance-based standards for air emissions under the Clean Air Act.32−36 The Clean Water Act greatly hindered the construction of new open-loop cooled power plants under §316(b), but grandfathered open-loop cooling at existing facilities. These open-loop power plants withdraw large volumes of water and return it directly to a water source, albeit at a higher temperature. These power plants often report zero water consumption via evaporation; however, the additional heat loading increases natural evaporation downstream.24 As an alternative, recirculating closed-loop cooling withdraws much lower volumes of water compared to open-loop cooling, but consumes over 60% of that volume.37 As open-loop power plants are retrofitted or replaced with closed-loop systems, these infrastructure changes might affect water availability and competition, motivating power plants to consider alternative water resources. Reclaimed water is poised as a viable alternative water resource for thermoelectric power plant cooling.38 Other Received: November 23, 2015 Revised: February 23, 2016 Accepted: April 14, 2016

A

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

time series, which allows us to bypass the previous assumption that streamflow and wastewater effluent are independent. Using the de facto quantification as a baseline, we compare the current (in situ de facto reuse) scenario with an engineered reuse scenario using reclaimed water in a piped network. Further, we include an intermediate recirculating cooling with de facto reuse scenario to identify the costs and benefits that are specific to engineered reuse. In our analysis, we focus on power plants as possible reclaimed water customers; however, our method is adaptable to include other consumers such as golf courses, agriculture, or other industrial cooling. Stillwell and Webber48 introduced a geospatial model based on least cost path analysis to evaluate the feasibility of cooling power plants with reclaimed water using a nonlinear optimization approach for individual power plants. The cost of a pipeline is a function of length, diameter, and a cost scaler that accounts for terrain variability. To determine the cost-scaling factor, we combine geospatial land use data with digital elevation models (DEM) of calculated slope into a raster geographic information systems format. The least cost path is then found between each wastewater treatment plant and each power plant. We advance the Stillwell and Webber48 method to include additional complexity by utilizing a genetic algorithm that considers all possible paths between wastewater treatment plants and power plants to select an optimal route. Practically, we employ the genetic algorithm associated with MATLAB; however, any genetic algorithm could be used. The cost of each pipeline is calculated using the same cost function from Stillwell and Webber48 with the flow, length, and cost scaler as inputs. The genetic algorithm treats the flow through each pipe as the decision variables, constraining flows as non-negative, and minimizes the sum of all the pipe costs. Results with zero flow represent a pipeline that would be suboptimal to construct. By formulating the optimization as a genetic algorithm, there is no need to assume any priority or water allocation rules. Although the solution would likely be expensive, this formulation allows for one power plant to be supplied by multiple wastewater treatment plants or one wastewater treatment plant to supply multiple power plants. The implications of transitioning from de facto reuse to engineered reuse and changing cooling technologies were evaluated using three feasibility metrics: cost, reliability, and performance. The cost of the generated reclaimed water network, are approximated using construction constants. Additionally, we estimate the cost to retrofit cooling towers at power plants, since recirculating cooling is generally necessary when using reclaimed water. We completed a first order approximation of cooling tower costs based on a previously published method for evaluating economic feasibility of cooling system retrofits at power plants.49 Since we are comparing only open- and closed-loop systems without accounting for economic value of drought resilience, the original formulation is adapted as follows:

common uses of reclaimed water include crop and landscape irrigation and dust control, with some areas of indirect potable reuse to augment drinking water supplies.39,40 The strengths of reclaimed water in power plant cooling applications include reliability and consistency of quantity and quality, without the environmental and legal risks of thermal pollution or entrainment and impingement issues for aquatic species. Reclaimed water also presents challenges of scale, corrosion, and biofouling, including danger of airborne bacteria that cause Legionnaire’s disease; however, these challenges can be minimized with proper planning and operations.41,42 Estimated costs for additional treatment of reclaimed water prior to use in cooling towers average $984/Mgal.47 Constructing any large infrastructure, such as a reclaimed water cooling system, can be expensive, prompting the need for well-informed decision-making. In the case of water and energy resources, spatial distribution often plays an important role. Multicriteria decision analysis tools have been developed to aid in this process.43,44 Our work builds on this vein of research by combining novel analytical approaches to quantitatively assess the suitability of using reclaimed water for power plant cooling. We employ scenario analysis to evaluate the implications of engineered water reuse compared to de facto reuse, defined by the incidental presence of wastewater in a water source. Originally developed to assess the percent of wastewater effluent at drinking water plants, we customize this de facto reuse method with temporal resolution to serve as a baseline in comparison to engineered reuse.45,46 Through our analysis, we present metrics of the system financial cost, reliability, and generation capacity to aid decision-making in support of sustainable energy and water resources management.



MATERIALS AND METHODS Since most power plants are downstream from municipal wastewater treatment plants, withdrawal of surface water sources for cooling leads to a degree of de facto water reuse. Quantifying the amount of de facto reuse establishes a baseline of existing hydrologic conditions at power generation facilities. In areas with significant levels of de facto reuse, the river channel acts as a natural conduit for transporting varying volumes of wastewater effluent. In such cases, construction of a reclaimed water distribution network might be less attractive than the de facto conditions since the end result would be similar without further investment. Quantifying de facto reuse is also important when considering quality. As described in literature, high levels of suspended solids and biologic material, present in wastewater, can cause fouling and other issues in cooling towers.38,42,47 In situations where de facto reuse is high, recirculating cooling might require additional treatment. In their work, Rice et al.45,46 quantified the percent of wastewater effluent present at a particular withdrawal point as % de facto reuse =

∑all i qw , i qs

⎡ i(1 + i)t ⎤ AC = P⎢ ⎥ + A O&M Q ⎣ (1 + i)t − 1 ⎦

(1)

where qw is the wastewater effluent from an upstream wastewater treatment plant i and qs is streamflow at the point of withdrawal, both in similar units. The Rice et al.45,46 analyses focused on drinking water treatment plants; we similarly quantify de facto reuse at power plants, changing only the withdrawal point. We expand on the existing method and use finer resolution daily data (when available) with correlating

(2)

where AC is the annualized cost [$/yr], P is the present value of the construction [$], AO&M is the annual operational cost per unit of cooling water [$/Mgal], Q is the annual reclaimed water flow rate [Mgal/yr], i is the annual interest rate, and t is the amortization period [years]. Literature values providing low and high estimates for constructing cooling towers, along with B

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

increases with the wet bulb air temperature, which in situations with high humidity the wet bulb temperature is assumed to be equal to the intake water temperature. For recirculating cooling, the inlet temperature is the sum of the source water temperature and the approach. For open-loop cooling, the inlet temperature is simply the source water temperature. Using historic averages of river temperature and wastewater effluent, we estimated the efficiency loss at each power plant (η), which directly relates the power plant capacity (N) with generation capacity loss (Nloss):

an estimated operations and maintenance cost, are provided in Supporting Information (SI) Table S1.50,51 The actual retrofit costs associated with cooling towers are site specific and include factors such as space requirements, geography, operations, and water quality; therefore, our analysis represents an initial approach to support water resources planning and decisionmaking. The costs associated with water treatment, fouling, and standard cooling tower maintenance are included to quantify operational expenses associated with reclaimed water.47 Beyond cost, reliability is an important metric to assess infrastructure. Reliability assesses the probability that a system is in a “satisfactory state” as defined by Hashimoto et al.52 Alternatively, reliability is defined mathematically, as shown in eq 3: Reliability = 1 − P[failure]

Nloss = ηN

(4)

In addition to any efficiency losses due to warm cooling water, we include parasitic pumping losses in the analysis of the reclaimed water (engineered reuse) scenario. Power plants are typically located next to a cooling water source such that cooling water pumping is approximated as negligible in the de facto scenario. Pumping large volumes of water considerable distances requires substantial energy to overcome changes in elevation, as well as major and minor friction losses in the distribution system. In the absence of a detailed pipe network, we account for only major losses due to friction (using the Hazen-Williams equation60; see the SI for details) and elevation changes. Operationally, we assume constant pumping rates over time, which is consistent with operations at baseload power plants.

(3)

where failure is defined as an unsatisfactory state. Data on detailed power plant operations (especially curtailments or efficiency losses) are scarce in public databases. To cope with this limitation, we define a failure as the number of days that a power plant requires a thermal variance. The U.S. Environmental Protection Agency (EPA) and state permitting agencies issue National Pollutant Discharge Elimination System (NPDES) permits to power plants regulating maximum cooling water effluent temperature(s), based on §316(a) of the Clean Water Act. On the occasion that a power plant cannot meet the temperature requirement, it is either fined or can request a temporary thermal variance from the state permitting agency. Although violations are public information,53 these data are often aggregated with no description of the cause of the fine. Consequently, we define a failure threshold as a granted thermal variance provision. Under the influence of policy and administrators’ discretion, this threshold might not be capturing many unsatisfactory states. A change in administration or policy could change the granting of thermal variance requests or ramifications for violating discharge temperature limits, or could require other actions altogether. Constructing cooling towers, regardless of water source, removes a large amount of the risk of environmental degradation or fines due to the small discharge rates. Failures are correlated with high temperatures. If the climate projections are accurate, failures will occur more frequently. We incorporate climatic warming by conditioning historical data on the percentage of temperature data that has exceeded the projected increase in average temperature. By assessing reliability, we aim to inform power plants and watershed managers in planning for both policy and hydrologic changes. With detailed power plant operational data, a more robust reliability analysis could be performed, accounting for environmental and regulatory risk. Finally, we quantitatively evaluated power generation performance when using reclaimed water for cooling. Intake water temperatures affect cooling systems, and can vary substantially in natural systems due to climatic factors and external forcings (e.g., upstream heat loading). Reclaimed water, on the other hand, is more consistent in temperature. We compare the efficiency losses for the three scenarios using a model introduced by Miara and Vörösmarty,54 which assumes an efficiency loss of 1.25% for every 1 kPa increase in the cooling system condenser pressure once a minimum threshold is reached. This minimum threshold is related to inlet temperature by the physical properties of water as a saturated liquid. For recirculating cooling, the condensing temperature



RESULTS To illustrate the proposed method, we analyze a watershed in northeastern Illinois, U.S., including the City of Chicago and several surrounding suburbs. Shown in Figure 1, the study area is comprised of three Hydrologic Unit Code (HUC) -8 basins that drain to form the Illinois River. The Illinois River and the waterways within our study area are important for barge transportation, connecting the Mississippi River to the Great Lakes. Of the river basins in the study area, Chicago is highly urbanized, the Des Plaines is suburban, and the Kankakee is primarily agricultural. The Chicago Area Waterways (CAW) are highly engineered, including several locks, dams, and diversions from Lake Michigan. Within the study area, there are 72 municipal wastewater treatment plants (WWTP) and six power plants. Cumulatively, the wastewater treatment plants discharge on average 1600 million gallons per day (MGD) during dry years, with three facilities (Stickney, North Side (O’Brian), and Calumet) managed by the Metropolitan Water Reclamation District of Greater Chicago (MWRD) contributing 80% of the total discharge flow. The study area power plants, which are described in Table 1, employ primarily open-loop cooling systems and use a variety of fuels. Our study area is situated such that the power plants are located near the mouth of the basin and downstream from many of the wastewater treatment plants. De facto Reuse. Using flow data from the gauging stations shown in SI Figure S1 and wastewater effluent averages, we calculate the median de facto reuse at each power plant. Although a straightforward calculation, the spatial aspects of the data are important. For our small urban watersheds, quantifying de facto reuse requires consideration of any discharges, withdrawals, or engineered operations of the waterways. In a few instances, discharges or withdrawals exist between the stream gauge and power plant. Figure 2 illustrates one of these instances (panel (B)) where a wastewater treatment plant might discharge downstream from a stream gauge. Under this C

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

events, stormwater combined with sanitary wastewater can overwhelm wastewater treatment infrastructure, causing a combined sewer overflow (CSO). Since wastewater bypasses the treatment plant (and, therefore, measurement), we do not have sufficient data to calculate de facto reuse during a CSO event; in response, we remove data associated with CSOs. We use the medians of the remaining data to calculate the de facto reuse at each power plant. The Will County power plant has the largest median de facto reuse at 65% while the two Joliet and Kendall County power plants are at 55% and 25%, respectively. (The two Joliet power plants are adjacent and therefore have the same de facto calculation.) The two nuclear power plants, Dresden and Braidwood, have de facto reuse less than 0.5%, due to withdrawals from the Kankakee River, a primarily agricultural basin that does not include large quantities of wastewater discharge. We can explain these results as a function of proximity to the large MWRD wastewater treatment plants. Following the waterway downstream, the de facto reuse percentage decreases because the catchment area contributes more streamflow while discharges from smaller wastewater treatment plants have minor effects. We analyze daily wastewater effluent and streamflow data from the MWRD and the U.S. Geological Survey (USGS), respectively, between the years 2007 and 2014. Daily data for the remaining wastewater treatment plants are unavailable; therefore, we approximate daily effluent flow from reported annual averages. In our study area, MWRD effluent comprises 85% of the total wastewater produced such that sufficient daily variation is captured. Upon first analysis, a large number of days yield a de facto reuse greater than 100%, which is inconsistent with the physical representation in eq 1. This finding reveals that on some days USGS stream gauges report less flow downstream than is reported being discharged from the wastewater treatment plants upstream. Our study area scale is sufficiently small to avoid time lag challenges; similarly, infiltration, evaporation, or unaccounted withdrawals do not appear to be of concern. We explain this result by the highly engineered and complex system of dams controlling the waterways. The modeling and research of these waterways is extensive, ongoing, and beyond the scope of our work.55−57 Therefore, we employ a one-week moving average to the data before calculating the de facto reuse. We represent the de facto reuse visually by depicting wastewater effluent (numerator in eq 1) against streamflow (denominator in eq 1), shown in Figure 3. Although the one-week moving average smoothing does not eliminate all the points greater than 100%, it reduced their number and magnitude. The

Figure 1. The Greater Chicago Area includes 72 wastewater treatment plants and six power plants. Of the power plants, five operate primarily by open loop cooling which cumulatively withdraw more water than the wastewater produced and are located on the downstream side of the study.

condition, we include the wastewater effluent in the numerator and denominator of the de facto calculation (using eq 1) since the upstream gauge does not account for its flow. We use similar mass balance logic for instances where two streams merge or the nearest gauge is downstream from the power plant. In the City of Chicago, as well as many older cities, the storm and sanitary sewers are combined, which is an important consideration in calculating de facto reuse. During large storm

Table 1. Study Area Power Plants Have Varying Characteristics and Costs to Retrofita water withdrawals name Will County Joliet 9 Joliet 29 Braidwood Dresden Kendall County

capacity (MW) 898 360 1320 2450 2020 1260

fuel

existing cooling system

coal coal coal nuclear nuclear natural gas

open-loop open-loop open-loopb open-loop open-loopb closed-loop

open loop (MGD) 607 263 956 1850 1440

retrofit to cooling towers

closed loop (MGD) 7.5 2.1 12.0 87.6 68.9 0.2

capital cost (106 US$) US$81 US$33 US$1,770

annual treatment costc (106 US$)

total annual cost (106 US$)

US$4.29 US$0.76 US$2.71 US$24.8 US$31.5 US$0.09

US$9.57 US$2.91 US$2.71 US$140 US$31.5 US$0.09

a

Results have been rounded. bHas facilities to operate as closed-loop but primarily utilizes open-loop cooling. cEstimated as $984 per million gallons.47 D

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 2. This hypothetical diagram illustrates the need to account for withdrawals and discharges that occur after the stream gauge and before the power plant in both the numerator and denominator of the de facto reuse calculation.

Figure 3. Conditioning to remove data that occurred on days with recorded combined sewer overflows, correlation exists between streamflow and wastewater effluent in the highly urban watershed of Chicago.

remaining percentages greater than 100, left of the dotted line in Figure 3, are within our margin of error. The regression plots in the left column of Figure 3

fact correlated due to the linear trend. Will County is the power plant nearest to the large wastewater treatment plants, which is reflected by the high slope of the trend line. The trend lines become flatter with increasing downstream distance, indicating

demonstrate that wastewater effluent and streamflow are in E

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology the location-specific nature of de facto reuse. These findings reveal that the assumption made by Rice et al.,45,46 that wastewater effluent is independent of streamflow is acceptable in most basins, but that assumption breaks down in highly urban environments, such as in the Chicago area. Representing the de facto reuse as a probability mass function in the right column of Figure 3, we find that de facto reuse varies substantially. As with the median de facto calculation, these probability mass functions reflect the proximity to the large MWRD wastewater treatment plants. The de facto reuse at Will County is wastewater dominated while Kendall County is runoff dominated. At Joliet 9 and 29 the de facto reuse is more distributed. Due to limited data availability for wastewater treatment plants in the Kankakee basin, no higher resolution analysis is performed. Since the two nuclear plants in this basin have such low preliminary de facto reuse percentages, a more precise analysis would likely reveal consistently low levels of de facto reuse. Engineered Reuse. To compare the de facto reuse scenario to an engineered reuse scenario, we formulate an optimal system to supply reclaimed water to power plants. Combining a digital elevation model and land use rasters from the USGS, we create a cost scaling raster for the greater Chicago area, shown as the background in SI Figure S3. Refining this method, we expand the cost scaling raster beyond the watershed boundary to allow the paths to traverse the least expensive route, with darker areas of SI Figure S3 indicating more expensive areas to build a pipeline. Topography in the study area is relatively flat, such that the cost scaling raster reflects differences in urban density. We simulate retrofitting power plants to use reclaimed water in recirculating cooling towers. Of the six power plants in the study area, only one (Kendall County) uses cooling towers; the remaining facilities operate open-loop systems, although Dresden and Joliet 29 have the necessary cooling towers on site. To determine the water withdrawal and consumption rates associated with retrofitting recirculating cooling, we use empirical and literature values specific to power generation in Illinois.11 Under this assumption of cooling system retrofits, the Stickney, North Side (O’Brian), and Calumet WWTPs each have enough effluent to supply all power plant demands in the study area. We find the least cost path between the wastewater treatment plants and power plants using the cost scaling raster with geographic information systems software (ArcMap by ESRI), displayed as the thin black lines in Figure 4. The genetic algorithm examines possible reclaimed water pipelines and selects the optimal solution, displayed as the thicker black line in Figure 4, representing piping reclaimed water from Stickney WWTP to each of the power plants. Cost. To approximate the cost of retrofitting power plants to use reclaimed water (engineered reuse), we used the average of the low and high estimates from SI Table S1 for each power plant in our study area, listed in Table 1. Due to the lack of data on the cost of cooling towers at nuclear power plants, there is high uncertainty in the retrofit cost estimate (see Stillwell and Webber48 for additional explanation). The estimated pipeline construction cost is $356 million, or $23 million/yr using a 30year amortization period and interest rate of 5%. The total length of pipe is estimated to be 93 miles long with diameters ranging from 0.5 to >6 ft. More details on section lengths and diameters are located in the SI. Similar feasible (yet

Figure 4. Least cost engineered reuse solution is a pipeline connecting the nearest treatment plant capable of providing all cooling demands.

suboptimal) solutions for complete sourcing from the Calumet or Northside (O’Brian) WWTPs reveal estimated costs of $423 million and $615 million, respectively. Combined, the total capital costs for the engineered reuse scenario is approximately $2.24 billion, with cooling tower costs representing 84% of the sum. This result is important when considering that the closed-loop cooling with de facto reuse scenario comprises the bulk of the capital costs required for engineered reuse. Naturally, de facto reuse with current cooling technologies, representing the baseline natural conditions, does not require any additional expense. These cost estimates represent a first-order approximation in motivating future indepth studies. Listed in Table 1, operation and maintenance costs for recirculating cooling, utilizing engineered reuse, are nonnegligible. These costs comprise about one-third of the total annual cost. This proportion is high due to the fouling and treatment costs associated with cooling with reclaimed water. We assume the operation and maintenance costs associated with open-loop cooling to be the baseline for comparison and, therefore, are zero. Due to lack of quality data, we cannot estimate treatment costs associated with recirculating cooling utilizing de facto reuse. However, due to the high presence of wastewater, we expect the costs to be closer to the engineered reuse than zero. F

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology Reliability. To calculate reliability, we quantitatively evaluate the likelihood of a power generation “failure” via a thermal variance event. We collect and organize documentation from the Illinois Environmental Protection Agency (IEPA) of thermal variances from 2003 to 2014.58 During this time period, 76 thermal variance days were recorded in the Chicago area out of 4015 total days. Using eq 3 and defining thermal variances as failures, we find the system of power plants in our study area is 98% reliable under de facto reuse conditions; however, this computation does not consider future climate shifts. We account for anticipated increases in streamflow temperatures (likely leading to additional thermal variance days) by conditioning the data on the 80th percentile of seasonal ambient air temperatures, leading to a simulated power generation reliability of 91%. We conditioned on the 80th percentile because it correlates to a modest 2.5 °F increase in the Chicago area average air temperature.59 We further grouped the variances into seasons for comparison to a seasonal climate metric, represented as the deviation from the seasonal average air temperature, illustrated in Figure 5. Most thermal variances occur during the drought of

2012; however, Dresden nuclear plant also had variances during 2005. Unlike the current de facto reuse conditions used to calculate reliability, reliance on engineered reuse introduces negligible power generation reliability concerns due to the relatively consistent quality and temperature of reclaimed water. The trade-off with a reclaimed water system is the reliance on critical pipeline infrastructure that is also at risk for failure, but leaving the existing cooling water intake structures as a backup can mitigate that risk. Performance. To assess the power plants’ operational performance under the de facto and engineered reuse scenarios, we model the capacity loss due to warmer cooling water and power consumed during reclaimed water pumping. Using reported average monthly intake temperatures from the Energy Information Administration for the years 2010 through 2013, we apply the capacity loss model (described in the Materials and Methods) for each of the power plants. Since we do not have detailed operational information on these power plants, we use estimates from literature for the threshold at which the intake temperature begins to affect capacity.54 Shown in Figure 6, the modeled capacity loss at each power plant is compiled (illustrated as stacked bars) to represent the total generation capacity loss for our study area. A peak capacity loss of 250 MW occurs for our de facto reuse scenario compared to a peak capacity loss of 400 MW for the engineered reuse scenario. The capacity loss under the de facto reuse scenario is due to the increased temperatures along the river, ranging from 26 to 29 °C. The maximum temperature of wastewater effluent, as reported by MWRD, is 23 °C, which is equal to the modeled threshold for efficiency loss in power plant cooling. Capacity loss in the engineered reuse scenario is the result of additional power demands for cooling tower operations. Although the engineered reuse scenario causes less capacity loss from elevated cooling water temperatures than the de facto scenario with recirculating cooling towers, we account for the pumping and distribution of reclaimed water from the wastewater treatment plant. (More details on the calculation can be found in the SI.) We find the power associated with pumping reclaimed water to the power plants is less than 1 MW. In comparing de facto reuse conditions with recirculating cooling (Figure 6(B)) and the engineered reuse scenario (Figure 6(C)), reclaimed water for power plant cooling is preferable due to substantially lower capacity losses due to consistency of water temperature, even when accounting for reclaimed water pumping. Notably, the capacity gains using reclaimed water, observed during summer months with peak

Figure 5. Without power plant operational data, we use thermal variances as a proxy for failure. Warmer seasons produce more thermal variances that have negative ramifications for the power plant and environment.

Figure 6. Capacity loss is greatest during the summer months due to the intake of warm cooling water. While recirculating cooling with cooling towers (B,C) introduces additional capacity loss compared to open-loop cooling (A), the impact is less pronounced in the engineered reuse case (C). G

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

discharges might encounter scrutiny and significant uncertainty regarding future operations. In such a future, engineered water reuse at power plants might be increasingly attractive from a policy perspective. An additional consideration regarding the suitability of engineered reuse is the price of the reclaimed water. Reported reclaimed water rates vary widely, typically ranging from $0.98 to $2.50 per 1000 gallons48 (some prices are up to $3.60 per 1000 gallons18), with many wastewater utilities setting rates less than drinking water prices to encourage use.18,63 While a complete economic analysis of reclaimed water pricing is beyond our scope, in our study area in Illinois, offering reclaimed water to power plants at little to no cost might incentivize cooling towers and protect against high thermal discharges in the summer. However, questions arise regarding ownership of wastewater effluent, especially in other geographic locations. In several western states, prior appropriation water rights laws account for return flows (wastewater discharges), such that downstream users might depend on wastewater effluent. Additional considerations such as power plant age, fuel sources, and air emissions constraints add to the uncertainty of future electric power generation. In some cases, utilities might be reluctant to invest in reclaimed water infrastructure for cooling of coal-fired power plants that are nearing the end of design life. In other circumstances, reclaimed water use might be advantageous due to siting flexibility and favorable economics. These factors motivate in-depth analyses using our proposed framework for site-specific evaluation of reclaimed water use for power plant cooling in the long term. These results suggest that although capital costs for reclaimed water infrastructure are high, the beneficial aspects of consistent water temperatures and reliability might outweigh these infrastructure investments. When considering factors such as the environment or public policy, cooling power plants with reclaimed water can minimize the risk of the unknown. Understanding and quantifying these benefits and trade-offs can help support sustainable water reuse and power generation.

electricity demand, are on the same scale as a small power plant. Using an electricity price of $0.08 per kWh61 and assuming the study area power plants would be operating at full capacity, we calculated a first order approximation of revenue loss of about $62 million/year due to cooling inefficiencies under engineered reuse compared to current de facto conditions. However, compared to de facto reuse with cooling towers, there is a net savings of $47 million/year, which exceeds the initial cost estimate for reclaimed water pipeline construction. That is, when retrofitting to use cooling towers, engineered reuse with reclaimed water provides economic advantages via improved performance. Overall, our results indicate that use of reclaimed water for power plant cooling has strategic advantages and trade-offs. The engineered reuse with recirculating cooling scenario reveals advantages compared to de facto baseline conditions in terms of reliability. These reliability gains are due to the predictable temperature of reclaimed water and its use in recirculating cooling towers, mitigating the need for thermal variances. When comparing recirculating cooling scenarios, engineered reuse with reclaimed water has lower capacity loss (that is, better performance) than recirculating cooling under de facto reuse conditions. These improvements in reliability and performance, however, come at the trade-off of increased infrastructure cost, yet estimated revenue loss from power plant derating is comparable to these investment costs. Consequently, use of reclaimed water for power plant cooling might be a strategic infrastructure investment to benefit both energy and water resources.



DISCUSSION AND POLICY IMPLICATIONS Our analysis reveals important insights regarding water reuse, both under existing de facto conditions and in a simulated engineered reuse scenario. Reclaimed water use for power plant cooling incurs high infrastructure investment costs for pipeline construction and cooling system retrofits. These investments, however, mitigate the risk of power generation reliability concerns from thermal variances and capacity loss due to elevated water temperatures when compared to cooling tower retrofits with existing water sources. The Clean Water Act developed the NPDES permitting system, which delegates authority to states to regulate the temperature of the cooling water discharged into waterways. In the state of Illinois via the IEPA, provisional thermal variances include exceptions for unreasonable hardship, a provision not included in the Clean Water Act or the NPDES permitting structure. In practice, the IEPA has accepted the claim that converting to closed-loop cooling would merit an excessive financial burden and presents a hardship. This conclusion favors the power plants; however, Micha62 argued that the policy might not be in accordance with federal laws. Provisional variances might violate the Clean Water Act because they essentially alter the water quality standards approved by the U.S. EPA.62 Understanding the gray area in which power plants receive thermal variances is important in assessing the suitability of reclaimed water. Retrofitting a power plant to incorporate cooling towers is expensive; however, reclaimed water use in conjunction with cooling towers has additional reliability and performance benefits, as demonstrated in our analysis. Considering the possibility of an intervention by the U.S. EPA, power plants might not be able to rely on the provisional thermal variances in the future. Similarly, NPDES permits often go through periodic review such that power plant cooling water



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b05753. Additional figures and methodology details are available (PDF)



AUTHOR INFORMATION

Corresponding Author

*Phone: 217-244-6507; fax: 217-333-0687; e-mail: ashlynn@ illinois.edu. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank undergraduate research assistant Lucas Djehdian, University of Illinois at Urbana-Champaign, for his support with data analysis and map creation. This work was supported by the Illinois Water Resources Center and the Department of Civil and Environmental Engineering. Author contributions: Z.A.B. completed the analysis and wrote the manuscript; A.S.S. formulated the study, supervised the analysis, and assisted with writing the manuscript. H

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology



(24) Badr, L.; Boardman, G.; Bigger, J. Review of Water Use in U.S. Thermoelectric Power Plants. J. Energy Eng. 2012, 138 (4), 246−257. (25) Roy, S. B.; Chen, L.; Girvetz, E. H.; Maurer, E. P.; Mills, W. B.; Grieb, T. M. Projecting water withdrawal and supply for future decades in the U.S. under climate change scenarios. Environ. Sci. Technol. 2012, 46 (5), 2545−2556. (26) Arpke, A.; Hutzler, N. Domestic Water Use in the United States A Life-Cycle Approach. J. Ind. Ecol. 2006, 10 (1), 169−184. (27) Ruddell, B.; Adams, E.; Rushforth, R.; Tidwell, V. C. Embedded resource accounting for coupled natural-human systems: An application to water resource impacts of the western U.S. electrical energy trade. Water Resour. Res. 2014, 50 (10), 7957−7972. (28) Fthenakis, V.; Kim, H. C. Life-cycle uses of water in U.S. electricity generation. Renewable Sustainable Energy Rev. 2010, 14 (7), 2039−2048. (29) Goldstein, R. A. A Survey of Water Use and Sustainability in the United States With a Focus on Power Generation; Electric Power Research Institute, 2003. (30) Kenny, J. F.; Barber, N. L.; Hutson, S. S.; Linsey, K. S.; Lovelace, J. K.; Maupin, M. a. Estimated Use of Water in the United States in 2005; U.S. Geological Survey, 2009; Vol. 1344. (31) Solley, W. B.; Pierce, R. R.; Perlman, H. A. Estimated Use of Water in the United States in 1995; U.S. Geological Survey, 1998. (32) Final Regulations To Establish Requirements for Cooling Water Intake Structures at Existing Facilities and Amend Requirements at Phase I Facilities; U.S. Environmental Protection Agency, 2014 (33) U.S. Army Corps of Engineers. Clean Water Rule: Definition of “‘Waters of the United States’”; U.S. Environmental Protection Agency, 2015. (34) U.S. Congress. Clean Water Act; 1972. (35) U.S. Environmental Protection Agency. Clean Power Plan; 2015 (36) U.S. Congress. The Clean Air Act; 2004. (37) Macknick, J.; Newmark, R.; Heath, G.; Hallett, K. C. Operational water consumption and withdrawal factors for electricity generating technologies: a review of existing literature. Environ. Res. Lett. 2012, 7 (4), 045802. (38) Dzombak, D. A.; Hsieh, M.; Li, H.; Chien, S.; Feng, Y. Reuse of Treated Internal or External Wastewaters in the Cooling Systems of Coal Based Thermoelectric Power Plants; National Energy Technology Laboratory, 2009. (39) Guidelines for Water Reuse; U.S. Environmental Protection Agency, 2012 (40) Water Reuse: Potential for Expanding the Nation’s Water Supply Through Reuse of Municipal Wastewater; National Research Council, 2012 (41) Li, H.; Chien, S.; Hsieh, M.; Dzombak, D. A.; Vidic, R. D. Escalating Water Demand for Energy Production and the Potential for Use of Treated Municipal Wastewater. Environ. Sci. Technol. 2011, 45 (10), 4195−4200. (42) Walker, M. E.; Safari, I.; Theregowda, R. B.; Hsieh, M. K.; Abbasian, J.; Arastoopour, H.; Dzombak, D. a.; Miller, D. C. Economic impact of condenser fouling in existing thermoelectric power plants. Energy 2012, 44 (1), 429−437. (43) Sattler, S.; Macknick, J.; Yates, D.; Flores-Lopez, F.; Lopez, a; Rogers, J. Linking electricity and water models to assess electricity choices at water-relevant scales. Environ. Res. Lett. 2012, 7 (4), 045804. (44) Stillwell, A. S.; Clayton, M. E.; Webber, M. E. Technical analysis of a river basin-based model of advanced power plant cooling technologies for mitigating water management challenges. Environ. Res. Lett. 2011, 6 (3), 034015. (45) Rice, J.; Wutich, A.; Westerhoff, P. Assessment of de facto wastewater reuse across the U.S.: Trends between 1980 and 2008. Environ. Sci. Technol. 2013, 47, 11099−11105. (46) Rice, J.; Westerhoff, P. Spatial and Temporal Variation in De Facto Wastewater Reuse in Drinking Water Systems across the U.S.A. Environ. Sci. Technol. 2014, 49 (2), 982−989. (47) Walker, M. E.; Theregowda, R. B.; Safari, I.; Abbasian, J.; Arastoopour, H.; Dzombak, D. A.; Hsieh, M. K.; Miller, D. C. Utilization of municipal wastewater for cooling in thermoelectric

REFERENCES

(1) Brown, T. C.; Foti, R.; Ramirez, J. a. Projected freshwater withdrawals in the United States under a changing climate. Water Resour. Res. 2013, 49 (3), 1259−1276. (2) Melillo, J. M.; Richmond, T. T. C.; Yohe, G. W. Climate Change Impacts in the United States: The Third National Climate Assessment, 2014. (3) Lubega, W. N.; Farid, A. M. Quantitative engineering systems modeling and analysis of the energy−water nexus. Appl. Energy 2014, 135, 142−157. (4) Sanders, K. T. Uncharted Waters? The Future of the ElectricityWater Nexus. Environ. Sci. Technol. 2015, 49 (1), 51−66. (5) Hardberger, A. Powering the tap dry: Regulatory alternatives for the energy-water nexus. Univ. Color. Law Rev. 2014, 84 (3), 530−579. (6) Malik, R. P. S. Water-Energy Nexus in Resource-poor Economies: The Indian Experience. Int. J. Water Resour. Dev. 2002, 18 (1), 47−58. (7) Bartos, M. D.; Chester, M. V. The conservation nexus: valuing interdependent water and energy savings in Arizona. Environ. Sci. Technol. 2014, 48 (4), 2139−2149. (8) Stillwell, A. S.; King, C. W.; Webber, M. E.; Duncan, I. J.; Hardberger, A. Energy-Water Nexus in Texas 2011, 16 (1), 2. (9) Santhosh, A.; Farid, A. M.; Youcef-Toumi, K. Real-time economic dispatch for the supply side of the energy-water nexus. Appl. Energy 2014, 122, 42−52. (10) Stillwell, A. S. Sustainability of Public Policy: Example from the Energy − Water Nexus. J. Water Resour. Plan. Manag. 2015, 141 (12), A4015001. (11) DeNooyer, T. A.; Peschel, J. M.; Zhang, Z.; Stillwell, A. S. Integrating water resources and power generation: The energy-water nexus in Illinois. Appl. Energy 2016, 162 (1), 363−371. (12) Stillwell, A. S.; Hoppock, D. C.; Webber, M. E. Energy Recovery from Wastewater Treatment Plants in the United States: A Case Study of the Energy-Water Nexus. Sustainability 2010, 2 (4), 945−962. (13) Stillwell, A. S.; King, C. W.; Webber, M. E. Desalination and Long-Haul Water Transfer as a Water Supply for Dallas, Texas: A Case Study of the Energy-Water Nexus in Texas. Texas Water J. 2010, 1 (1), 33−41. (14) Tidwell, V. C.; Kobos, P. H.; Malczynski, L. A.; Klise, G.; Castillo, C. R. Exploring the Water-Thermoelectric Power Nexus. J. Water Resour. Plan. Manag. 2012, 138 (5), 491−501. (15) Scanlon, B. R.; Duncan, I.; Reedy, R. C. Drought and the water−energy nexus in Texas. Environ. Res. Lett. 2013, 8 (4), 045033. (16) Frumhoff, P. C.; Burkett, V.; Jackson, R. B.; Newmark, R.; Overpeck, J.; Webber, M. Vulnerabilities and opportunities at the nexus of electricity, water and climate. Environ. Res. Lett. 2015, 10 (8), 080201. (17) Chen, L.; Roy, S. B.; Goldstein, R. a. Projected Freshwater Withdrawals Under Efficiency Scenarios for Electricity Generation and Municipal Use in the United States for 2030 1. J. Am. Water Resour. Assoc. 2013, 49 (1), 231−246. (18) Barker, Z. A.; Stillwell, A. S.; Berglund, E. Z. Energy and water tradeoffs in the expansion of a dual water system. In revision. (19) Feeley, T. J.; Skone, T. J.; Stiegel, G. J.; McNemar, A.; Nemeth, M.; Schimmoller, B.; Murphy, J. T.; Manfredo, L. Water: A critical resource in the thermoelectric power industry. Energy 2008, 33 (1), 1−11. (20) Chandel, M. K.; Pratson, L. F.; Jackson, R. B. The potential impacts of climate-change policy on freshwater use in thermoelectric power generation. Energy Policy 2011, 39 (10), 6234−6242. (21) Tidwell, V. C.; Macknick, J.; Zemlick, K.; Sanchez, J.; Woldeyesus, T. Transitioning to zero freshwater withdrawal in the U.S. for thermoelectric generation. Appl. Energy 2014, 131, 508−516. (22) Shuster, E. Estimating Freshwater Needs to Meet Future Thermoelectric Generation Requirements; National Energy Technology Laboratory, 2007. (23) Förster, H.; Lilliestam, J. Modeling thermoelectric power generation in view of climate change. Reg. Environ. Chang. 2010, 10 (4), 327−338. I

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology power plants: Evaluation of the combined cost of makeup water treatment and increased condenser fouling. Energy 2013, 60 (February), 139−147. (48) Stillwell, A. S.; Webber, M. E. Geographic, technologic, and economic analysis of using reclaimed water for thermoelectric power plant cooling. Environ. Sci. Technol. 2014, 48 (8), 4588−4595. (49) Stillwell, A. S.; Webber, M. E. Novel methodology for evaluating economic feasibility of low-water cooling technology retrofits at power plants. Water Policy 2013, 15 (2), 292. (50) FPLE - Beacon Solar Energy Project Dry Cooling Evaluation; WorleyParsons: Folsom, CA, 2008. (51) Zhai, H.; Rubin, E. S. Performance and cost of wet and dry cooling systems for pulverized coal power plants with and without carbon capture and storage. Energy Policy 2010, 38 (10), 5653−5660. (52) Hashimoto, T.; Loucks, D. P.; Stedinger, J. R. Robustness of water resources systems. Water Resour. Res. 1982, 18 (1), 21−26. (53) U.S. Environmental Protection Agency. Enforcement and Compliance History Online https://echo.epa.gov/. (54) Miara, A.; Vörösmarty, C. J. A dynamic model to assess tradeoffs in power production and riverine ecosystem protection. Environ. Sci. Process. Impacts 2013, 15 (6), 1113−1126. (55) Waterman, D. M.; Waratuke, A. R.; Motta, D.; Cataño-Lopera, Y. a.; Zhang, H.; García, M. H. In Situ Characterization of Resuspended-Sediment Oxygen Demand in Bubbly Creek, Chicago, Illinois. J. Environ. Eng. 2011, 137 (8), 717−730. (56) Jackson, P. R.; García, C. M.; Oberg, K. A.; Johnson, K. K.; García, M. H. Density currents in the Chicago River: Characterization, effects on water quality, and potential sources. Sci. Total Environ. 2008, 401 (1−3), 130−143. (57) Cantone, J.; Schmidt, A. Improved understanding and prediction of the hydrologic response of highly urbanized catchments through development of the Illinois Urban Hydrologic Model. Water Resour. Res. 2011, 47 (8), W08538. (58) State of Illinois. Illinois Government News Network http:// www.illinois.gov/news (accessed July 28, 2015). (59) Chicago Metropolitan Agency for Planning. Appendix A: Primary Impacts of Climate Change in the Chicago Region; Chicago, 2013. (60) Mays, L. W. Water Resources Engineering, 2nd ed.; John Wiley & Sons, Inc: Hoboken, NJ, 2011. (61) U.S. Energy Information Administration. State Electricity Profiles. http://www.eia.gov/electricity/state/. (62) Micha, P. K. In hot water: Clean Water Act provisional variances and their relationship to the impact of heat waves and droughts on the supply and demand of electricity. Chicago-Kent J. Environ. Energy Law 2014, 4 (1). (63) Carpenter, G.; Grinnell, G.; Haney, C.; Jacobi, G.; Koorn, S.; O’Reilly, D.; Pierce, C.; Riley, C.; Rimer, A.; Thompson, K.; et al. Water Reuse Rates and Charges 2000 and 2007 Survey Results, 2008.

J

DOI: 10.1021/acs.est.5b05753 Environ. Sci. Technol. XXXX, XXX, XXX−XXX