This article was originally published in the November/December 1999 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.


| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |



Home Energy Magazine Online November/December 1999

Money Down the Drain: 
Controlling Hot Water 
Recirculation Costs

by Fredric S. Goldner

Fredric S. Goldner, a Certified Energy Manager, is founding principal of Energy Management & Research Associates and adjunct professor at the Center for Energy Policy and Research at the New York Institute of Technology. He can be reached at

Domestic hot water recirculation systems provide steady streams of hot water to top-floor tenants in multifamily buildings--but at a significant energy cost. This energy penalty can be slashed by 40% with the use of a simple return line aquastat, while still keeping the occupants happy.

The heat computer is the boiler controller that the author adapted to collect data on DHW consumption, recirculation flow, DHW temperature before and after the mixing valve, boiler run time, fuel consumption, and recirculation pump run times. 
For control strategy D, the aquastat was set to 110°F with a 5°F deadband.
Figure 1. Strategy D--operating the recirculation pumps only when the DHW temperature in the return line falls below 110°F--shaved fuel use by almost 11%.
Figure 2. With recirculation pumps running continuously, DHW recirculation systems eat up 38% of the fuel used to produce DHW.

Three Types of DHW Piping Systems

Most multifamily buildings have one of three types of DHW return/recirculation system. The first system has no recirculation piping at all. This type of system is most often found in the smallest buildings, where there is a short run between the supply source (boiler or heater) and the farthest tap. The second is a gravity recirculation system. Heat always rises; so when the hot water gets heated in a boiler, even if no one is using it, the hot water will rise through the pipes, and as it rises it will push the cooler water around. The rate at which it does so is very slow, usually less than 0.5 gpm. The third option is a forced recirculation system. These systems employ a small pump to keep water flowing, thus avoiding stagnation and the need for residents to run the tap for a long time (particularly on upper floors) to get hot water. The pumps are run continuously or are cycled on and off by an aquastat or a timer. Most buildings with aquastats have them set at 180°F, which means that the pumps run whenever the temperature of the DHW in the return line falls below 180°F; essentially these pumps run continuously.
Peter Spinillo, research assistant for Energy Management & Research Associates, adjusts the aquastat setting. The aquastat, which sits on top of a DHW return pipe, controls the recirculation pump. (The aquastat is wired into the DHW recirc pump behind him.) 
Energy use for domestic hot water (DHW) is the second largest component of a multifamily building's energy budget; it is surpassed only by heating in cold and mixed climates. To provide DHW that reaches an acceptable temperature at the tap without having to wait too long, and to reduce the amount of water that runs down the drain while residents wait for warm water, multifamily building developers commonly install some type of DHW recirculation system (see Three Types of DHW Piping Systems). DHW recirculation pumps push hot water throughout the building, often in uninsulated pipes, so that it is readily available even to tenants on the top floors--an effective, but energy-intensive solution (see Controlling Recirculation Loop Heat Losses, HE Jan/Feb '93, p. 9). General practice is to run the recirculation pumps continuously. But is this necessary? Or can DHW system energy consumption be reduced by applying a different control strategy to the operation of the pumps?

I recently completed a study, funded by the New York State Energy Research and Development Authority (NYSERDA), that analyzed the energy savings and effects on water temperature and availability that could be attained by varying control strategies of DHW recirculation systems in multifamily buildings. Six buildings were included in the study, with two sites in each of three size ranges: small (fewer than 45 apartments), medium (45 to 80 apartments), and large (more than 80 apartments). Before the study began, all six buildings had forced recirculation systems, with the pumps running continuously.

To measure the energy savings achievable by cycling the pumps, I compared four operating strategies:

  • Strategy A was to operate the pumps continuously. Strategy A was used as a base case against which to determine the savings achievable by the other strategies.
  • Strategy B was to try to save energy by shutting down the pumps at night, from 11:50 pm to 5:20 am, when few if any people use hot water.
  • Strategy C was to try to save energy by shutting down the pumps during peak morning and evening DHW usage periods--from 5:45 am to 8:15 am and again from 5:45 pm to 9:15 pm. During these periods, the large volume of water being used keeps hot water in circulation without additional pumping.
  • Strategy D was to operate the pumps only when the DHW temperature in the basement return line fell below 110°F. Pump cycling is controlled by a return line aquastat--a type of thermostat for water--set at 110°F with a 5°F deadband.
Instrumented monitoring collected data on five-minute DHW and recirculation flows; five-minute running average of DHW temperatures measured after the mixing valve and in the return line; and daily circulation pump run times. It was essential that the analysis compare the effects of each strategy at each site in order to isolate the effects that were attributable directly to the change in recirculation pump operations. To that end, the data for each building were compared under each of the four strategies to account for differences in use patterns that exist between buildings. To alleviate seasonal DHW consumption effects, the sites were operated for two weeks under each of the four strategies, and the entire eight-week testing round was conducted four times, once during each of the four seasons. This round-robin approach made it possible to collect a full two months of data for each strategy, and it eliminated the problem of seasonal variations that would have arisen if each strategy had been tested over one eight-week period.

The number of persons living in the building also greatly affects DHW consumption. Working with the building management, who keep excellent records, I used vacancy figures to eliminate DHW consumption variances caused by changing occupancy levels, and to adjust the consumption of DHW per occupied apartment. Adjustments were also made to account for changes in occupant behavior in response to variable inlet water temperatures encountered in each period.

Tenant Satisfaction To learn how the tenants responded to variations in DHW delivery caused by the four different recirculation control strategies, I developed a series of questionnaires. Each tenant was surveyed after each strategy, but in a different testing round. (I and my assistant, Peter Spinillo, also regularly interviewed the building superintendents to see if there had been complaints during the testing.) The tenants were asked to answer four questions related to whether they had to wait for hot water, how acceptable the temperature of the water was, and whether their water use had changed during the past two weeks.

Responses showed that the level of water delivery and the temperature of the hot water were satisfactory under each of the strategies, except in the case of Strategy C. With Strategy C, about 27% of respondents felt that they always had to wait for hot water and about 7% felt that the water that arrived was never hot. In contrast, with Strategy A, during which the second highest percentage of respondents registered complaints, not quite 15% of the tenants waited for hot water and for roughly 3% of them the water was never hot. The level of satisfaction with Strategy D was similar to that with Strategy A.

Energy Use Compared to the Strategy A base case, Strategies B and C each saved 5.5% and Strategy D saved 10.8% in fuel oil used (see Figure 1). The savings were the greatest for Strategy D because with this strategy the annual average pump run turned out to be approximately six hours per day, although the actual hours varied from season to season (see The Whole Picture).

These are meaningful savings. The energy savings from Strategy D is equal to a savings of approximately 4% of a typical building's total annual fuel consumption. On average, 38% of the total water-heating energy used to supply DHW during normal operations (base case Strategy A) was expended solely to heat the recirculated DHW (see Figure 2). The research results show that the energy spent on heating recirculating water was cut to an average of 31% of the DHW system total by Strategy B, to 34% by Strategy C, and down to 23% by Strategy D.

In addition to the fuel savings, the reduced run hours on the pumps yielded electrical energy savings. Under Strategy D, these savings ranged from $5 to $13 per year in buildings with a 1/4 hp pump, and from $16 to $43 per year in buildings with a 1/12 hp pump; these amounts will vary depending on the applicable electric rate.

The short pump run time was somewhat surprising, but it can be explained by two facts. First, pump cycling was unnecessary during peak hours, and second, the water's heat loss was reduced because the water was not being forcibly and continuously recirculated around the building. DHW leaves the mixing valves at a temperature of 130°F-140°F. The less time the DHW is running through the pipes, the longer it takes for the hot water to drop the 20°F-30°F needed to trigger recirculation.

Water Consumption and Temperatures One of the major concerns before beginning the research was that, with the recirculation pump turned off, people would let the water run for a long time in order to get hot water. However, there was no appreciable difference in the volume of city water used among any of the strategies. The volume of DHW being recirculated under nonbase case strategies was greatly reduced, which accounted for most of the energy savings under these strategies. Typically there was a moderate drop off in recirculated water--approximately 25%--from Strategy A to Strategies B and C, and a much more significant drop-off with Strategy D. Switching from Strategy A to Strategy D reduced the amount of water recirculated by an average of 74% annually.

To see if different water temperatures were contributing to the energy savings, actual temperatures measured throughout the system were used in the calculation of the energy usage during each test period. The reduction in DHW temperature at the basement return under Strategies B, C, and D did contribute to the energy savings, as the systems were not circulating higher-temperature DHW around the buildings as often under these three strategies.

Fast Payback Significant energy savings can result from installing a rather simple, off-the-shelf device--the aquastat. Payback can be achieved in just a few months. Fuel savings from Strategy D in the study sites averaged $780 per year per site. (Savings ranged from $117 to $1,784, depending on the size of the building.) As I explained above, electrical energy savings ranged from $5 to $43 per year; these savings are attributable to the reduced run hours on the pumps. The aquastats used in this project cost $43, and were acquired from a general supply house. Conservative installed cost estimates from several heating and plumbing contractors for this measure range between $150 and $250, and, where local codes allow, the measure could be installed by building staff. The payback period from this straightforward operations change ranged from just over one month to one and one-half years; the average payback was seven months. A reverse acting return line aquastat control should be installed immediately in any building that has a continuously running forced recirculation system.

Further Reading

Domestic Water Heating Design Manual. Westlake Village, California: American Society of Plumbing Engineers, 1998.

The Whole Picture

Figure S-1.
Figure S-2.
A comparison of Figures S-1 and S-2 (Building 7 during Round 2, Scenario A and Scenario D, weekdays) provides an excellent overall illustration of what is happening in the buildings' DHW systems. Here we see the DHW consumption and recirculation flows plotted against both the supply and the return water temperatures. 

When you look at Figure S-1, you may ask is it really necessary to circulate 130°F-145°F DHW around the building's piping system? When you look at Figure S-2, you will see that the answer to this question is No. Note that the supply temperature of the hot water available at any given five-minute time period during the day is almost identical in the two graphs. 

Further review of these two figures reveals changes in both the recirculation consumption curve and volume, as well as the temperature of the water in the return piping. The energy savings comes from the enormous reduction during overnight and late-night periods, along with some smaller but significant savings during all but the highest early morning and evening peak consumption periods.



 | Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |

Home Energy can be reached at:
Home Energy magazine -- Please read our Copyright Notice



  • 1
  • NEXT
  • LAST
SPONSORED CONTENT What is Home Performance? Learn about the largest association dedicated to home performance and weatherization contractors. Learn more! Watch Video