1、NA-04-5- 1 Hot Water Use in a High School Cafeteria Carl C. Hiller, Ph.D., P.E. Member ASHRAE Jeffrey Miller David R. Dinse, P.E. Member ASHRAE ABSTRACT Little information is available on how hot water is used in schools. This paper reports feld test data on hot water and water heating energy use of
2、 a high school cafeteria kitchen. Objectives were to obtain hot water and energy use informa- tion and to investigate potential energy eficiency improve- ments both analytically and experimentally. To achieve these objectives, a suitable test site was located and the cafeteria kitchen gas-fired hot
3、water recirculation loop was insru- mented and monitored. Performance of the system was moni- tored for several months in the as-found condition, with the hot water recirculation-loop pump running continuously. Subse- quently, timer control of the recirculation-loop pump was implemented in order to
4、reduce energy consumption, and monitoring was continued for the remainder of the 12-month test period. The configurations were then analyzed as if each had been operating for the entire 12 months, using the actual temperatures and amounts of hot water consumed as observed each day. Several additiona
5、l improved configurations were also investigated analytically. Energy consumption, overall delivered eficiency, and energy savings were computed on a 12-month basis for the various configurations and the results presented here. Some of the results presented are believed representative ofmost high sc
6、hool kitchen operations, most notably the fact that schools are in use only about 50 % ofthe days of each year and normally for less than half of each day when they are in use. This has important generalized implications for appropriate school water heating system design and operation, as discussed
7、in the papel: INTRODUCTION Information on potable hot water use in schools is gener- ally lacking, as is documented data on hot water recirculation- loop piping heat loss. One can use piping heat loss calculation methods available in most heat transfer textbooks or the ASHRAE handbooks to estimate h
8、ot water recirculation-loop system piping heat loss. Such calculations show that, in most applications, heat loss from recirculation-loop piping is large if the loop is even moderately long, even if the pipes are well insulated. This loss can be reduced somewhat if the recircula- tion-loop pump is o
9、perated only when end uses require hot water, but losses are typically still significant. This paper describes field test results on the cafeteria kitchen water heat- ing system in a high school in Portland, Tennessee. Objectives were to quantitatively determine hot water use, quanti recir- culation
10、-loop heat loss, and to determine system annual energy use under several water heating system design and control strategies. Several system design and control improve- ment options have been investigated either experimentally or analytically as part of this work. Both generalized and site- specific
11、findings, conclusions, and recommendations are discussed in this paper. METHODOLOGY A suitable hot water recirculation-loop system test site was located, instrumented, and monitored in the as-found configuration for a period of time (Hiller and Miller 2002) The system was then modified by the additi
12、on of time-clock control of the recirculation-loop pump and monitored frther. The total monitoring period was 12 months. Performance of the two configurations and several other theoretical design and control options for the system was then compared by analyz- Carl C. Hiller is president of Applied E
13、nergy Technology Company, Davis, Calif, Jeffrey Miller is a senior engineer at AIL Research, Inc., Princeton, N.J. David R. Dime is a project manager at Tennessee Valley Authority, Chattanooga, Tenn. 02004 ASHRAE. 655 ing the performance of each configuration for the full 12- month period, using act
14、ual observed hot water consumption and temperatures for each day. To do this, the effective heat loss characteristics (UA value) of the tanks and plumbing were determined from experimental data and then applied to the observed temperatures and water consumption. It was neces- sary to fill in for sev
15、eral different periods of lost hot water use or energy use data. This was possible because temperature data were obtained consistently throughout the test period, as were data from another water heating system at the school, making it possible to identify non-occupied periods. Test Site Selection Nu
16、merous schools were examined before the final test site was selected. Portland High School in Portland, Tennes- see, was selected as the test site for the following reasons: The school has eight separate hot water recirculation- loop systems. This allowed monitoring of two fairly small loops, making
17、 the project feasible within budget- ary constraints. A separate ASHRAE paper has already been presented, describing results on the other water heating loop (Hiller et al. 2002). The school was fairly new (opened in fall 1997), and the water heating systems were of typical design and well implemente
18、d, using typical modem practice. School board officials and school personnel were inter- ested in energy efficiency and were willing to cooperate with test personnel. Test Site Characteristics The school had a total enrollment of approximately 950 students. The kitchen served both breakfast and lunc
19、h, with many more students eating lunch than breakfast. The exact number of meals served was unknown and generally variable from day to day. The cafeteria kitchen was served by two gas-fired tank- type water heaters operating in parallel. They were located in a small equipment room immediately adjac
20、ent to the kitchen. A relatively short hot water recirculation-loop hot water distri- bution system was installed to serve the fixtures listed below. Distance to the farthest fixture from the water heaters was approximately 75 feet. However, it was not possible to measure exact recirculation-loop di
21、mensions and distances because much of the piping was under the floor slab, and the rest was overhead in a sealed-access area due to fire code requirements for the kitchen. The fixtures on the kitchen circuit were: lavatory (bathroom) sink (1) foot-petal-operated hand sinks (4) clothes washer (for w
22、ashing towels, washcloths, aprons, etc.) (1) commercial dishwasher (with its own 10 kW electric resistance booster water heater) (1) janitorial sink (1) food disposal sink (1) large food preparationipot and pan washing sinks (4) reel-type spray hoses (for washing sinks, countertops, floors) (7) food
23、 serving counter warming trays (2) Characteristics of each of the two water heaters (plumbed in parallel) were: natural gas fired external finned-tube heat exchanger with pumped circu- lation between heat exchanger and storage tank nominal 100 gal (379 L) storage tank burner input rating 199,000 Btu
24、k (58 kW thermal), yielding a tank full reheat time of approximately 30 min manufacturer claimed nominal input efficiency 82% 1/6 hp (0.124 kW) nominal water circulating pump motor output rating The recirculation-loop water pump had a nominal 1/12 hp (0.062 kW) motor output rating. Both the recircul
25、ation-loop pump and the water heater pumps were running continuously in the as-found configuration. Moreover, it was determined that while the recirculation-loop pump could be shut off when not needed, continuous running of the water heater circulating pumps was required for proper operation of the
26、water heaters. Analysis of gas use data suggests that water Bow was rela- tively equally distributed between the two water heaters, with both water heaters firing approximately equal amounts of time. School maintenance personnel had adjusted flow control valves on both water heaters to help achieve
27、this result. Instrumentation and Data Collection collecting the following data: A single central data acquisition system was utilized, Thot, Tcold, Tair and water flow at the water heaters gas consumption at the gas water heaters kWh of the recirculating-loop pump Data were stored at one-minute inte
28、rvals and downloaded daily over telephone lines to a central data-processing loca- tion. On-site data storage was provided for up to ten days in case of phone line or modem difficulties. In addition, a one-time measurement was taken of the power consumed by the two gas water heater circulating pumps
29、. These pumps, which ran continuously, used approxi- mately 191 watts each. By comparison, the hot water recircu- lation-loop pump used an average of approximately 86 watts when running. A single gas meter was used to monitor gas use to both water heaters. Magnitude of gas use showed when one versus
30、 two water heaters was firing. A single water meter was also used. The data collection period was from January 20, 2001, through January 19,2002. Monitoring of the as-found uncon- 656 ASHRAE Transactions: Symposia trolled hot water recirculation-loop (RLU) system ran from January 20, 2001, through A
31、pril 3, 2001. Monitoring of the original system, but with timer control on the hot water recir- culation-loop pump (RLC), ran from April 3, 2001, through the end of the test on January 19,2002. The recirculation-loop pump was turned off from approximately 11:OO p.m. to 5:30 a.m., or approximately 6.
32、5 hours daily. Further analysis after the control strategy was set indicated that the pump could have (and should have) been shut off at 2:OO p.m., allowing a shut- down period of 15.5 hours (see the discussion that follows). Data Fill and Annual Analysis Procedures The water flow meter on the kitch
33、en circuit was not installed until April 6,2001, due to difficulties in shutting off the main water supply valve to the school. Once the valve was successfully turned off, a secondary shutoff valve was installed to the line serving the kitchen so that it could be isolated in the future. Moreover, th
34、e kitchen flow meter devel- oped a water leak that damaged the meters pulse counting electronics, and readings stopped on October 17, 2001. This means that kitchen water flow data were collected for around 6.5 months. In order to perform full year energy consumption analysis and comparisons, a final
35、 kitchen water flow data set was created by filling in missing data with data from other days for which data were obtained. Data from the temperature sensors and the other water loop separately monitored at the school (Hiller et al. 2002) were used to determine days when the school was closed and, h
36、ence, when the kitchen had no flow. The flow data indicated that water use in the kitchen was somewhat different before versus after the summer break, with average daily use somewhat lower in the fall than in the spring. This was probably due to changes in kitchen personnel. For the unoccupied days
37、of the missing flow data period (approxi- mately 37% of the missing flow data period), no error was introduced by the data fill procedure. For the remaining 63% of the data fill days, an unknown uncertainty was introduced by the data fill procedure. Temperature data indicated that no hot water short
38、ages occurred during the missing flow data period, and filled-in data reflected this fact. In order to estimate annual energy use under a variety of system operating strategies, heat loss UA factors were deter- mined separately for the tanks and recirculation-loop piping. Total energy use was then e
39、stimated by separately estimating tank heat loss, piping heat loss, and energy use to heat deliv- ered hot water, using assumed burner heat input efficiencies. From data taken on days when no hot water was used, it was possible to separately identi the standby gas consump- tion of the gas water heat
40、er recirculation-loop combination. Furthermore, using additional information from the timer- controlled test period, standby gas consumption could be separately identified for just the gas water heater. Subtracting, the standby gas consumption for keeping the recirculation- loop hot could also be de
41、termined. The gas meter was installed and was operational from January 20, 2001, until March 1, 2001, a period of only 38 days. The gas meter was removed by school personnel on March 1,2001, in the mistaken belief that it was restricting gas flow, resulting in recurrent hot water shortages in the ki
42、tchen (despite test personnel assurances to the contrary). The cause of the kitchen hot water shortages was later determined to be a faulty float-actuated fill valve on the commercial dish- washer, which allowed hot water to overflow continuously to drain whenever the dishwasher was in operation. Th
43、e fill valve was repaired on March 9, 2001, and the hot water shortage problem was resolved. Several attempts were made to have plumbers reinstall the gas meter, but the plumbers commis- sioned to do the work repeatedly failed to do the work as planned. Note that the gas meter and water flow meter w
44、ere never installed and operational at the same time. Fortunately, it is periods of zero flow that allow determination of tank and recir- culation-loop heat loss characteristics, and data from the other monitored loop indicated days when the school was unoccu- pied (and, hence, when there was no wat
45、er consumption in the kitchen loop) while the gas meter was operational. Since energy used to actually heat the water can be reliably deter- mined by knowing the entering cold water temperature, deliv- ery temperature, amount of water used, and heat input device efficiency, total energy use could be
46、 estimated for all days of the test, This was done by summing the energy used to make up tank and recirculation-loop heat loss and the energy that was used to actually heat the delivered hot water. It was possible to determine standby energy use of the tank plus recirculation-loop combination direct
47、ly from the zero-draw-day data collected in February when the gas meter was installed. It was also possible to identify the amount of gas used by the standing pilot lights on both water heaters. Addi- tionally, it was possible to develop a correlation between the number of temperature rises on the t
48、ank outlet water pipe under zero flow conditions and the number of main burner firings and an estimate of the average amount of gas used during each main burner firing when in standby mode. It was thus possible to determine the amount of gas use in standby mode (zero draw) when the recirculation-loo
49、p was turned off by using the correlation between outlet temperature and main burner gas use developed from the February data. By subtrac- tion, it was then possible to determine the gas use in standby forjust the piping. This allowed determination of effective UA factors for the tanks and recirculation-loop piping separately. Heat loss was then estimated for each day of the year by apply- ing the observed average hot water to equipment room air temperature differential to the tank and piping UA values. RESULTS AND DISCUSSION Hot Water Use Results The cafeteria served both breakfast and
