1、4714 (RP-1172) Domestic Hot Water End Use Analysis Methods and Preliminary Results Dale K. Tiller, D.Phil. Ronny Goepfert ABSTRACT A new analysis method to allocate domestic hot water draws to specific end uses has been developed under the auspices of ASHRAE project I 172-RF The “temperature- based
2、event inference” method allocates hot water use events based on measured water flow at the hot water tank and changes in end use pipe temperature. The change in pipe temperature is an indirect measure used to allocate waterflow to speclJc end uses. This paper describes the methods, proce- dures, and
3、 selected preliminary results obtained by applying this method to a sample of four residences in Omaha, Nebraska. We also present recommendations for future work. INTRODUCTION ASHRAE has reported several initiatives related to domestic hot water (DHW) use in residences, condominiums, and commercial
4、spaces (e.g., Abrams 1998; Abrams and Shedd 1996; Becker and Stogsdill1990a, 1990bl; Becker et al. 199 i; Fanney 19901; Hiller 19981; Lowenstein and Hiller 1996, 19981; Perlman and Mills 1985; Perlman and Milligan 19881). This work has provided the technical foun- dation for current ASHRAE recommend
5、ations (ASHRAE 1999). Additional work has been conducted by the Electric Power Research Institute (Ladd and Harrison 1985), the Gas Research Institute (Paul et al. 1994), the American Water- works Association in collaboration with municipal utilities (DeOreo et al. 1996a, 1996b, 2001; DeOreo and May
6、er 1996; Mayer and DeOreo 1993, and the United States Environmen- tal Protection Agency in collaboration with municipal utilities (Mayer et al. 2000, 2003). Two methods have been developed to characterize DHW use: the flow trace signature analysis method and the temper- Gregor P. Henze, Ph.D., P.E.
7、Member ASHRAE Xin Guo ature-based event inference method. The flow trace signature analysis method is well documented (e.g., DeOreo et al. 1996al; Hiller 19981; Lowenstein and Hiller 1996, 19981) and has been applied in studies of residential water conserva- tion programs (e.g., Mayer et al. 2003).
8、This method employs flow measurements at the hot water tank and a few supporting temperature measurements at the main branches of the hot water pipe network. Data acquisition systems are there- fore unobtrusive and relatively quick to install and commis- sion. Extensive test runs at individual hot w
9、ater end uses are required to characterize flow rates and event durations, as well as significant analysis program customization to account for different flow rates and event durations observed at individual sites. It is not always possible with this method to differentiate between end uses that exh
10、ibit similar flow rate and duration profiles, and it is sometimes difficult to disagregate concurrent draws. Lowenstein and Hiller (1998) discussed the difficulty in discriminating concurrent draws using the flow trace signature method. They estimated that between 4% and 15% of hot water draws repre
11、sented concurrent flows. They also noted that human interruption of automatic dish- and clothes-wash- ing machine cycles and complex control cycles on automatic machines can make it difficult to correctly attribute hot water draws based on flow measurements collected at the hot water tank only. Lowe
12、nstein and Hiller (1998) supplemented flow measurements from the hot water tank with temperature measurements of branch hot water pipe feeds. Although this technique helped them more accurately attribute water flow, they noted that more measurement points located at the hot water draw location itsel
13、f would have provided more specific localization of hot water use. Dale K. Tiller and Gregor P. Henze are associate professors, Ronny Goepfert was an international research student, and Xin Guo is a grad- uate student in the Architectural Engineering Department, University of Nebraska-Lincoln, Omaha
14、, Neb. 21 8 02004 ASHRAE Under the auspices of ASHRAE project 1 172-RP, Henze et al. (2002) developed new analysis methods and software tools that allocate domestic hot water draws to specific end uses, which they called the ?temperature-based event infer- ence method.? As in the flow trace signatur
15、e method, a flow- meter is used to measure flow directly at the hot water tank. Measured flow is allocated to specific end uses and end use location based on the sharp increase in pipe temperature that occurs at the end use when hot water flows through the pipe. Each end use (dishwasher, clothes was
16、her, water tap, etc.) is wired with a fast-responding temperature sensor and sampled at a suficiently high frequency so that no event ?slips through? undetected. The change in pipe temperature provides an indi- rect measure that is used to allocate water flow to specific end uses. The temperature-ba
17、sed event inference method requires more on-site commissioning and decommissioning, but it offers fully automated and accurate end use resolution of single and multiple hot water draws, with little investigator input once the hot water pipe network in the house has been instrumented with thermocoupl
18、es. Henze et al. (2002) reported results from a month-long pilot study of a single residence that compared flow trace signature analysis and the temperature-based event inference method. The temperature-based event inference method was able to correctly allocate 97.1 % of the hot water draw events t
19、o different end uses, while flow trace signature analysis was able to correctly allocate 90.6% of hot water draws to different end uses. This paper describes the methods, procedures, and selected results obtained by applying the temperature-based event inference method to a sample of four residences
20、 in Omaha, Nebraska. We also present recommendations for future work. EX P E R I M EN TAL SET UP Description of Studied Residences Hot water use was studied in four residences located in metropolitan Omaha, Nebraska, over the fall of 2001 and winter of 2002. These residences differed in age, occupan
21、cy, and the number of hot water end uses. One house (Elk Creek neighborhood) was a two-story detached house occupied by one adult. The house has a finished basement, which is equipped with a full bathroom (toilet, sink, and shower). The hot water tank (rated capacity 50 U.S. gallons, natural gas fue
22、l) is located in a separate basement utility room. The data acqui- sition system was located in this utility room. The ground floor has a half-bath, equipped with a toilet and a sink. The kitchen and laundry room are also located on the ground floor. The kitchen is equipped with a double sink and a
23、dishwasher. The laundry room has a washing machine and a dryer. Three bedrooms are located on the second floor. The master bedroom has an ensuite bathroom, equipped with a toilet, two sinks, a large soaker tub, and separate shower stall. A second bathroom on the second floor is equipped with a toile
24、t, sink, and a combined bathtubshower. The second house studied (Country Club neighborhood) was a two-story detached house, occupied by two adults and two young (five years old) children. The house has a finished basement, with a separate laundry room. The hot water tank (rated capacity 40 U.S. gall
25、ons, natural gas fuel) is located in the basement laundry room. The data acquisition system was located in this utility room, concealed to the side of the furnace, which prevented access by children. The laundry room is equipped with a washing machine, a laundry tub, and a dryer. The kitchen is loca
26、ted on the ground floor and is equipped with a sink and a dishwasher. The bedrooms and bathrooms are located on the second floor. The main bathroom is equipped with a toilet, a single sink, and a combined bath- tublshower. The second bedroom has an ensuite bathroom and is equipped with a toilet, one
27、 sink, and a combined bathtub/ shower. The third house included in the study (Fairacres neigh- borhood) was a two-story row house, occupied by two adults and two dogs. The house has a finished basement, with a sepa- rate laundry room. The hot water tank (rated capacity 40 U.S. gallons, natural gas f
28、uel) is located in a separate basement util- ity room. The data acquisition system was located in this util- ity room. The laundry room is equipped with a washing machine and a dryer. A separate half-bath is located in the basement, which is equipped with a shower stall, and a sink. The kitchen is l
29、ocated on the ground floor and is equipped with a double sink and a dishwasher. The ground floor also has a half-bath, equipped with a toilet and a sink. The master bedroom is located on the second floor and has an ensuite bath, which has a toilet, a single sink, and a combined bathtub- shower. The
30、fourth residence studied was a two-bedroom apart- ment (Bellevue neighbourhood) occupied by one adult and one young child ( five years old). The apartment has its own hot water tank (rated capacity 40 U.S. gallons, natural gas fuel), which is located in a small utility closet off the kitchen area. T
31、he data acquisition system was located outside the util- ity closet, concealed behind and under a table. This apartment has a separate laundry room, which is equipped with a washing machine and a dryer. The kitchen is equipped with a double sink and a dishwasher. The apartment has one bathroom, whic
32、h is equipped with a toilet, a single sink, and a combined bathtubshower. Data Acquisition and Management Overview Figure 1 depicts an overview of the data acquisition management and control system architecture installed in each residence. Type-T thermocouples were attached to each hot water end use
33、 in each house. These wires were passed back through the residence to a personal computer (PC)-based data acquisition management and control system. The data acqui- sition interface to the PC is a multifunction universal serial bus (USB) data acquisition module that offers a wide dynamic range, auto
34、 ranging, and thermocouple support. The modules used feature 16 single-ended or 8 differential 16-bit A/D ASHRAE Transactions: Research 21 9 Measurement points (hot water pipes ) USE Hub T Pers on al Com puter , 1 FIOW meter I Figure 1 Data acquisition system architecture. inputs, with a maximum sam
35、pling rate of 50 kS/s. Software selectable gain settings of 1, 10, 100, or 500 provide input ranges of k10 V, k1 V, kO.l V, and h20 mV. With a range of *10 V, a resolution of 16 bit, and a gain of 500, the smallest voltage change that can be detected by the AD converter is 0.6 pV, which should be su
36、fficient to record small temperature changes. The device features software enabled cold junction compensation. We selected this device because the USB interface offers easy and convenient connec- tion to the PC. Because the USE3 modules reside outside the PC, they are susceptible to ground spikes, w
37、hich could cause a systems crash. The module features 500 V galvanic isolation that protects the PC from ground spikes and ensures a reliable stream of data. In each residence, a flowmeter was attached to the water pipe leaving the hot water tank. The signal from the flowmeter was also connected to
38、the data acquisition system. The PC and data acquisition system were plugged into the electrical power supply via an uninterruptible power supply that provides four minutes of backup power to the system in the event of power failure. Although power outages of long duration are infre- quent in Omaha,
39、 Nebraska, short-term outages of less than a second in duration are relatively common. The data acquisition system polled each measurement point every second, resulting in about 18 megabytes of data each day. These raw data files were automatically processed on a daily basis, using a custom software
40、 application that iden- tified and extracted individual draw events from the larger data files (more details on the analysis are provided below). This analysis resulted in a much smaller text file (less than 1 O kilobytes), describing individual hot water draws. These hot water end use data files we
41、re transferred on a daily basis from the residence to a central location for further analysis and eventual uploading to an Internet-accessible database using an onboard 56K modem that was included with the PC data acquisition system. During the course of the study, two modems failed, and these were
42、replaced with external modems. Each of the following sections provides more detail on important features of data collection, analysis, and manage- ment. End Use Pipe Temperature Measurement and Thermocouple Placement As noted above, the temperature-based event inference method allocates hot water dr
43、aws based on temperature change at the end use pipe. Type-T thermocouples were installed at every end use and at two branches in the pipe network leading to the upper floors in each residence. Care was taken to ensure that the thermocouples were mounted as close as possible to the device or service
44、actually using the water (e.g., sink or dishwasher) because thermal conduction in the pipe can make it difficult to discriminate between hot water draws that branch from a common pipe (e.g., as often occurs at kitchen sinks and dishwashers). The thermocouple tip was 220 ASHRAE Transactions: Research
45、 fastened to pipes using two separate layers of electrical tape, and, if necessary, a plastic tie wrap was also used to more firmly affix the thermocouple in place. Two layers of electrical tape were used to avoid inadvert- ently creating a tertiary copper junction that could have affected thermocou
46、ple performance. Before the thermocouple was fastened to the pipe, a layer of electrical tape was wound firmly around the pipe. The thermocouple tip was then placed next to the tape layer covering the pipe, and then a second layer of electrical tape was wound around the pipe and thermocou- ple tip.
47、This proved a very effective method of mounting ther- mocouples to end use pipes. All thermocouples remained in place for the duration of the study, with only one exception. In this case, the thermocouple was attached to a bathtub faucet that was located directly under a leaky shower head. Water dro
48、pping from the shower head onto the faucet eventually weakened the tape seal, causing the thermocouple to detach from the faucet. Plastic tie wraps were used in addition to the electrical tape, and the thermocouple remained firmly attached to the faucet. Thermocouples connected to the basement and g
49、round floor end use points were passed through wall and floor cavi- ties and were then strung across the basement ceiling (inside aplenum, ifavailable) to the computer data acquisition station. The ends of each thermocouple wire were connected to the differential input channels of the data acquisition units. Thermocouples connected to upper floor end uses were bundled together and concealed inside plastic conduit. In resi- dence 1, plastic conduit was used to conceal the wire bundle all the way back to the basement ceiling plenum. The wire bundle from the upstairs l
