Dissecting Rainwater Pump Energy Use in Urban Households Tjandraatmadja, G. 1, Pollard, C. 1, Sharma, A. 1 and Gardner, T. 2 1 CSIRO Land and Water, PO Box 56, Highett,Vic 3190 2 CSIRO Land and Water, Ecosciences Precinct, Dutton Park, Qld 4102 Summary Rainwater tanks are a viable alternative to reduce the demand of mains water and increase the resilience of cities to drought. However, previous studies have found that the energy footprint of rainwater systems can be much higher than the energy footprint for traditional water distribution and is also subject to high variability across various systems in use. This study investigated the operation of pumps used for rainwater supply in a controlled residential environment (a model house) to understand the factors affecting the energy footprint for rainwater supply in urban areas. Pumps commonly installed in urban households operate more efficiently for high flow applications (>15 L/min). However, the majority of household water end uses requires low water flow (<10 L/min), causing the pump to operate in the low energy efficiency range. Energy efficiency varies for each pump. Correct pump sizing and informed selection of system components could improve energy efficiency significantly to <1.5 kwh/kl, lower than or close to the energy footprint of other alternative water sources such as desalination and recycled water. Keywords Rainwater tanks, energy, pump system configuration. Introduction The widespread uptake of rainwater tanks in urban areas has provided a viable alternative to reduce the demand of mains water and increase the resilience of cities to drought. On the other hand, studies on energy consumption indicate that the energy footprint of rainwater systems, expressed in kilowatt hours per kilolitre (kwh/kl), can be much higher than the energy footprint for traditional water distribution systems (Gardner et al., 2006; Beal et al., 2008; Retamal et al., 2009; SEWL, 2009). Field studies also show that there is significant variability in energy requirements in rainwater supply associated with the various system configurations, technologies, and operation regimens adopted in urban areas, whilst limited information is available on how to optimise their performance (Beal et al., 2008; Retamal et al., 2009; SEWL, 2010; Talebpour et al., 2011). This study investigated the water flow characteristics and the operation of pumps used for rainwater supply in a controlled residential environment (a model house) to understand the factors impacting on the energy footprint for rainwater supply in urban areas. Methodology In the model house rainwater from a tank was supplied by an external pump through ¾ inch polyethylene pipe to the range of common household appliances (Figure 1 and Table 1). Pressure and flow within the water supply system and energy consumed by the pump were monitored in real time at intervals of 0.2s and logged every 1s. To reduce water consumption, water was recycled within the model house. Figure 1. Diagram of rainwater supply system in model house
Table 1. Characteristics of appliances adopted in model house. Appliance Average water consumption WELS * Classification Washing machine (5.5kg capacity) 113L/wash 4 stars Dishwasher 15.7L/wash 2.5 stars Toilet cistern Dual flush (6L/3L) n.d. Tap Up to 35L/min n.d. Note: * Water Efficiency Labelling Scheme (Australian Government 2011) Findings Rainwater End Uses and Water Needs On remote properties, rainwater is often the sole supply of water adopted for all potable and non-potable end uses. Whilst in urban areas, rainwater is used mainly for non-potable applications, such as outdoor uses, supply of washing machine cold water and filling of toilet cisterns. Water requirements for household end uses are curtailed within certain limits dictated by design constraints and by the pressure in water supply system. Water consumption of household appliances and flow rates of fittings, such as taps, showerheads and irrigation devices, are curtailed to increase water efficiency (Australian Government, 2011). In addition, household appliances, such as washing machines and dishwashers are designed to operate within a specific range of flow and/or pressure conditions. Table 2 shows the typical design pressure and flow rates for fittings and some common household appliances sold in Australia. For instance, a washing machine s minimum operating pressure can range from 40 to 100 KPa depending on appliance make and model, and it can sustain pressures up to 1 MPa. Whilst flow from a water efficient (four to six star Water Efficient Labelling Scheme (WELS)) tap can be restricted from 2 to 7 L/min (Australian Government, 2011). Table 2. Design parameters for appliance operation Appliance Minimum Pressure (kpa) Maximum Pressure (kpa) Source Washing machine 40-100 800-1000 Products technical specifications Dishwasher 30-150 800-1000 Products technical specifications Toilet cistern 150 * 400* Standards Australia (1999) Tap (WELS 4-6 stars) n.a. 2-7L/min Australian Government (2011) Note: * Standards Australia performance test conditions for water inlet. In addition, individual end uses are characterised by a typical pattern of water supply. A washing machine and a dishwasher typically require multiple water supply events during a wash cycle, which will require a pump to start multiple times. Whilst hand washing and cistern filling tend to be a single continuous event of short duration. Pump Operation Earlier studies on tank water supply to dwellings were conducted with a range of pumps with motor power ranging from 0.35 kw to 1.25 kw (Beal et al., 2008; Gardner et al., 2008; Retamal et al., 2009; SEWL, 2009 and 2010). In these studies, the majority of the pumps had motor capacities within the ranges of 0.41 to 0.6 kw (44%) and 0.61 to 0.81 kw (33%). In this study, three fixed speed external pumps were examined. Pumps A, B and C had respective capacities of 0.20, 0.55 and 0.75 kw power and could deliver 31, 33 and 45 m maximum head. Fixed speed pumps are the most common type of pump adopted for rainwater systems in urban dwellings (Retamal et al., 2009). Pump selection was conducted to include the two most common pump sizes and a smaller size pump. Pumps A and C were triggered by pressure switches and pump B by a mains water switch. After all pumps were tested, the rig was later modified to incorporate pressure vessels. This paper will incorporate results for pump C and an 18 L pressure vessel as an example. The energy footprint for rainwater supply for each pump was evaluated for a range of water supply conditions including the operation of each appliance investigated. For any pump, as flow increases the specific energy required for delivery of water decreases, reaching an optimal (minimum) specific energy at higher flow rates (>15 L/min), as shown in the pump curves in Figure 2.
The appliances and end uses tested in this study were toilet cistern filling, hand washing, dishwasher and a washing machine., The operating flow rate requirements during operation were 3.3-4.1 L/min (the dishwasher), 4-6.3 L/min (toilet cistern filling and hand washing) and 9.6-13 L/min (washing machine). Figure 2 shows the range of raintank water flows and the average specific energy measured for water provision by each pump to the washing machine, toilet cistern filling and dishwasher using the three pump sizes in the model house (grey boxes). Figure 2. Energy requirements for supply of rainwater to household appliances. The mean specific energy requirements for each end use ranged from 0.6 to 5.3 kwh/kl for individual end uses and the 3 different size pumps. For instance, pump A supplies rainwater to the washing machine at 10+0.23 L/min and specific energy 0.6+0.05 kwh/kl. Typical end uses such as toilet, hand washing and dishwashing resulted in high specific energy requirements because of the low flow associated with their operation. The highest flow rate observed was for the washing machine. Thus for a same pump, a rainwater tank connected solely to a toilet cistern would be more energy intensive than a rainwater tank connected solely to a washing machine. Figure 2 also shows that energy differed across the three pump sizes, with the difference particularly exacerbated in the low flow and high energy use range of the curve (<10 L/min in Figure 2). However, as flow is increased to more than 20 L/min all three pumps displayed optimum performance (optimal energy efficiency). Therefore, reduction of the specific energy for water supply requires either an increase in the water flow requirements during pump operation or adoption of a pump with a low energy versus flow curve. Energy Distribution Pressure losses due to friction within the system were minimal (less than 6%) for all three pumps. The majority of the energy was used for tank water transfer. Energy distribution during a pump operation event is comprised of 3 basic stages: (a) Pump start-up; (b) Flow; and (c) Over run (Retamal et al., 2009). Figure 3 illustrates the typical pump operating profile.
Figure 3: Typical pump operating profile. As illustrated in Figure 4 for toilet cistern filling, for the larger capacity 0.55 and 0.75 kw pumps, the energy required for pump start-up was minimal (<6% total energy use). However, for the smaller pump (the 0.20 kw pump A), a larger proportion of the energy was consumed in the start-up of low flow end uses. For example, a full toilet cistern filling start-up consumed 46% of the energy as multiple stop-starts occurred. Energy for the over-run stage also varied across pumps, with negligible energy dissipated in pump A and the most in pump C (25% of total event energy for the half flush). Figure 4. Energy requirements for supply of rainwater to household toilet cistern. Pump Comparison Overall, comparison of the specific energy requirements for the three pumps tested showed that the specific energy consumption (Figure 5) and the overall energy consumption for each end use (Figure 6) were lower for the smaller pump, pump A. The variation in specific energy was particularly evident for low flow events such as the dishwasher and toilet cistern filling. However, for end uses with a higher flow of water delivery, such as the washing machine, the differences in specific energy consumption became less marked between pumps, as expected from observation of Figure 2. Hence in this case, choosing the smaller pump A over pump C could reduce the energy for toilet cistern filling by 2/3 (from 3.5kWh/kL to less than 1.5kWh/kL). However, pump design (brand) also needs to be considered in the selection process. Pumps A and B (same manufacturer) produced a very similar specific energy curve despite their power difference (Figure 2). Complementary studies comparing eight different pumps have also shown that the specific energy consumption could vary significantly between pumps of same capacity but of different brands (Hauber-Davidson and Shott 2011, SEWL 2010).
Figure 5. Specific energy requirements for operation of common household appliances. Figure 6. Total energy consumption for pump water supply. The pressure provided during pumps operation decreased with pump capacity, but all the three pumps tested were able to deliver more than sufficient pressure to operate the appliances tested (Figure 7). The system configuration adopted (rainwater tank + fixed speed pump + appliance) is the typical set-up adopted for rainwater supply in urban dwellings. These findings confirm some of the initial hypotheses developed through observation of in-situ households by Retamal et al., (2009), including that significant energy savings can be achieved by matching end uses needs and pump operation. Retamal et al. (2009) had also concluded that low energy intensity for a rainwater system was not necessarily correlated with lower overall energy consumption in a household. Our outcomes indicate that, for the system adopted, a smaller size pump is able to deliver a satisfactory level of service and use less energy overall for the supply of individual indoor appliances in a same system. The results also show that the energy consumption is not highly influenced by the number of start-ups from a pump, but more often by the flow rate required for water supply. Figure 7. Average pressure provided during pump operation. The Role of Pressure Vessels Retamal et al. (2009) suspected from the analysis of household energy consumption that pressure vessels could deliver potential energy savings, but were not able to quantify the actual savings. Hauber-Davidson and Shott (2011) verified that a small pressure vessel (5 L capacity) had no effect on pump energy use. Our results indicate that energy savings can be gained when appropriate system configuration, pressure vessel size and end use requirements are considered. The volume of water that a pressure vessel will hold is dictated by the vessel capacity and pump operating pressure. Typically, a pressure vessel holds less water than their nominal volume. Energy savings are generated if the volume of water that the pressure vessel provides is equal to or greater than the total volume of water required for an end use. For instance, this can be illustrated for pump C coupled with an 18 L pressure vessel. An 18 L pressure vessel connected to pump C holds 6.3+0.09 L, and was able to provide for low volume uses (<6 L), such as hand washing, toilet cistern filling and dishwasher operation as well as intermediate water supply during a washing machine cycle. As a result, the pump started less often, but more importantly, when the pump filled the pressure vessel it operated at a higher flow rate (average 9.5+0.3 L/min). This led to more efficient pump operation as illustrated in Figure 8, particularly for low volume and low flow end uses.
Figure 8. Pump C (0.75 kw) energy consumption with and without an 18 L pressure vessel (PV). Conclusions Pumps commonly installed in urban households for rainwater supply operate more efficiently for high flow applications (>15 L/min). However, typical end uses for mandated rainwater supply, such as toilet flushing and washing machine cold water tap, often operate at flows much less than 15 L/min. This in turn causes pumps to operate in their low energy efficiency range resulting in a high specific energy compared with traditional potable water supply, e.g. 0.68 kwh/kl for Brisbane (Kenway et al., 2008). Better matching end uses needs and pump size can improve energy efficiency significantly to less than 1.5 kwh/kl, resulting in a lower energy footprint than sea water desalination (3.6 kwh/kl (SEWL, 2009)) or Indirect Potable Reuse (2.8-3.8 kwh/kl (in NSW, 2006)), the common alternatives to supplement dam sourced potable water. Other devices such as pressures vessels can also contribute to reduce energy requirements if adequately sized. References Australian Government (2011) WELS Products, http://www.waterrating.gov.au/products/index.html, updated May 2011, accessed May 2011. Beal, CD, Hood, B, Gardner, T, Lane, J and Christiansen, C (2008). Energy and water metabolism of a sustainable subdivision in South East Queensland: a reality check, Enviro 2008, Melbourne. Gardner, E.A., Millar, G.E., Christiansen, C., Vieritz, A.M. and Chapman, H. (2006). Energy and water use at a WSUD subdivision in Brisbane, Australia, Australian Journal of Water Resources, 10( 3). Hauber-Davidson, G. and Shott, J (2011) Energy consumption of domestic rainwater tanks why supplying rainwater uses more energy than it should, Water,38(3),72-76. Kenway, S. (2008). Preliminary LCA of the SEQ Water Strategy, Urban Water Security Research Alliance, September 2008. NSW (2006). A sustainable water supply for Sydney, New South Wales Legislative Council, Sydney. SEWL (2009). Energy consumption in domestic rainwater harvesting, Prepared by Water Conservation Group for South East Water Limited, South East Water Limited. SEWL (2010). Energy consumption of domestic rainwater harvesting, Prepared by Water Conservation Group for South East Water Limited, May 2010. Standards Australia (1999) AS 1172.2 Water closet (WC) pans of 6/3L capacity or proven equivalent- Part 2: cistern, Standards Australia. Retamal, M, Glassmire, J, Abeysuriya, K, Turner, A, and White, S. (2009). The Water-Energy Nexus: Investigation into the Energy Implications of Household Rainwater Systems, CSIRO-Institute for Sustainable Futures, University of Technology, Sydney. Talebpour, MR, Stewart, RA, Beal, C, Dowling, B, Sharma, A and Fane, S (2011). Rainwater energy tank end usage and energy demand: a pilot study, Water, March 2011, 97-101.