Here is the text version of the webinar, "Central Multifamily Water Heating Systems," presented January 21, 2015.

Elizabeth Weitzel, Alliance for Residential Building Innovation (ARBI)
Jordan Dentz, Advanced Residential Integrated Energy Solutions (ARIES)
Eric Ansanelli, Advanced Residential Integrated Energy Solutions (ARIES)


Gail Werren:
Hello, everyone, I'm Gail Werren with the National Renewable Energy Laboratory and I'd like to welcome you to today's webinar hosted by the Building America program. We are excited to have Elizabeth Weitzel, Jordan Dentz, and Eric Ansanelli joining us today to talk about effective use of central heat pump water heaters and control systems to reduce the energy use in hot water distributions.

Before we begin, I’ll quickly go over some of the webinar features. For audio, you have two options. You may either listen through your computer or telephone. If you choose to listen through your computer, please select the “mic and speakers” option in the audio pane. By doing so, we will eliminate the possibility of feedback and echo.

If you select the telephone option, a box on the right side will display the telephone number and audio PIN you should use to dial in. If you have technical difficulties with the webinar, you may contact the GoToWebinar’s Help Desk at 888.259.3826, for assistance. If you would like to ask a question, please use the “Questions” pane to type in your question and, if possible, please indicate to whom it is addressed.

If you are having trouble viewing the materials through the webinar portal, you may find PDF copies of the presentations at the website listed here and you may follow along as our speakers present. Today’s webinar is being recorded and the recording will be available on the DOE YouTube channel within a few weeks. We have an exciting program prepared for you today that will focus on the effective use of central heat pump water heaters and control systems to reduce the energy use in hot water distributions.

Before our speakers begin, I will provide a short overview of the Building America program. Following the presentations, we will have a question and answer session, closing remarks, and a brief survey. The U.S. Department of Energy’s Building America program has been a source of innovations in residential building energy performance, durability, quality, affordability, and comfort for nearly 20 years. This world-class research program partners with industry to bring cutting-edge innovations and resources to market.

Building America is supported by 10 industry research teams and four national labs. Each of these teams, and labs, partner with dozens of industry professionals including builders, remodelers, manufacturers, and utilities. The best and the brightest in the residential buildings industry can be found here.

Building America uses applied research to deliver building science solutions, using a four-step framework. These innovative solutions are tested in homes to develop proven case studies of success the market can point to. Building America provides the tools the building industry needs to ensure the innovations are applied correctly, always keeping an eye on energy performance, durability, quality and affordability. The final step, infrastructure development, is the conduit to getting innovations to the marketplace.

Building America research focuses on how the components of new and existing homes work together through systems integration. As the market changes and evolves, so has the direction of our research in order to add value and drive changes in performance across the building industry. In addition to technical challenges we have been addressing for decades, there is now a need to understand market transformation issues, such as valuation of energy efficiency. In the 20 years of Building America research, we have spearheaded combining ultra-high efficiency with high performance in both new and existing homes. And we are consistently achieving this challenging task.

For example, in 1995, a typical home used three times more energy per square foot compared to today, and indoor air quality, comfort, and durability problems were common. Today, a home built to DOE Zero Energy Ready Home specifications uses less than half the energy and is more comfortable, healthy, and durable. By 2030, Building America will demonstrate that new and existing homes can produce more energy than they use.

Do you want to know more about these proven innovations? The Building America Solution Center is your one-stop source for expert information on hundreds of high-performance construction topics, including air sealing and insulation, HVAC components, windows, indoor air quality, and much more. You can find it by the URL on your screen or by searching on "Building America Solution Center." Also, the Building America website provides information about the program, the latest Top Innovations and case studies, and there, you can also subscribe to the monthly newsletter. And now, onto today’s presentation.

Our webinar today will provide information about improving efficiency of central heat pump water heaters and control strategies that can reduce the energy use of hot water distribution systems. If you would like detailed information about any of these efforts, or if you are interested in collaborating, please feel free to contact any of our presenters.

Our first speaker today is Beth Weitzel, who is a senior engineer at Davis Energy Group, the lead for the Building America Team Alliance for Residential Building Innovation. Today Beth will share the results for an evaluation of a 10.5-ton central heat pump water heater installed at the UC–Davis West Village Zero Net Energy Community, which is one of 45 similar units installed in student apartment buildings at West Village. The heat pump water heater was monitored from October 2011 through February 2013 and provided useful information to the developer and installation contractors as they worked through early issues in implementing a technology new to them.

Our next presenters are Jordan Dentz and Eric Ansanelli from the Levy Partnership, the lead organization for the Building America team, Advanced Residential Integrated Energy Solutions. Jordan is a vice-president at the Levy Partnership where he develops and manages building science research projects and consults for building owners and developers. Eric is a mechanical engineer with expertise in steam heating, specializing in energy auditing, modeling and consulting. Jordan and Eric will summarize the results of a field study in which two types of central domestic hot water controls, demand control and temperature modulation, were retrofitted into four existing multi-family dwellings in New York. They will describe energy and cost savings and resident satisfaction, as well as interactive space conditioning systems.

With that, I'd like to welcome Beth to start the presentation.

Elizabeth Weitzel:
Thank-you. So today I will be sharing the results of a field evaluation of a 10 1/2 ton heat pump water heater that was installed at the University of California - Davis' West Village student housing. But before I go any further I just want to add the disclaimer that the results presented are not representative of certified, rated, testing conditions, that they are field performance data, and the conditions and operation is different than those used for performance standards. So with that in mind, we will continue on.

In the background, I'll lightly go over the history and reasons for central heat pump water heaters, the site project, and system configuration. Our objectives for the project were to measure field performance and reliability of a central heat pump water heater -- most notably how often the heat pump water heater meets the load and under what conditions does it need backup. We then wanted to compare the economics of running the heat pump water heater system to conventional electric storage and natural gas storage options. I'll go over the observations we made during the monitoring period and how we took the monitoring data and developed a heat pump water heater model in transit to compare the economic comparisons.    

So, why heat pump water heaters? According to the Department of Energy more than 40% of U.S. households use electric resistance water heating. There have been extensive studies on single-family homes that have demonstrated that heat pump water heaters can provide 50% or more energy savings compared to electric resistance. There's been less attention on multifamily heat pump water heating, however, as the product options are more limited and at the time of this study there have not been many central heat pump water heaters installed and monitored. Most heat pump ... most multifamily units typically use a central storage or boiler system or have distributed water heating.

So the project install that we had the opportunity to monitor was at the University of California Davis at a site called West Village, which is a mixed use commercial and student housing complex. The site is located in Davis, California, outside of Sacramento, which is a specified hot-dry climate. While in California natural gas service is more typical, West Village is a targeted Zero Net Energy community. It was planned to have enough PV to supplement all electric service to the site, as well as having advanced items like a bio-gas system that worked on agricultural waste.

There are 45 student apartment buildings, each with 12 units. The building that we monitored had 36 bedrooms, but there's about 2,000 individual bedrooms on site. The buildings are 14,200 square foot and served by a 10.5 ton heat pump water heater, with electric resistance backup. The backup consists of two 120 gallon electric resistance heaters that are 54 kilowatts each. The building was also served by a recirculation pump that was both temperature and timer controlled.

So this is a diagram, system diagram, of our installation. Cold water, you can see my mouse here, enters the building from the service and mixes with the recirculation loop and branches to both the heat pump water heater and the first storage unit. The heat pump water heater is a direct heat exchange with the first storage tank. The supply from the first tank then goes into the secondary tank and then finally to the building supply after a mixing valve. Our sensors were placed to capture the total heat pump energy delivered, domestic hot water energy delivered, recirculation losses, and total energy electric usage.

Heather:
Beth, this is Heather. Sorry to interrupt. You have your control panel up. Thank-you. You fixed it now. We need the control panel closed to see the whole thing. Thank-you so much.

Elizabeth Weitzel:
Oh, sorry. Our period of monitoring began in October 2011 and continued until January 2013, so that we could collect at least a full year’s data after the initial commissioning phase.

So while monitoring was intended for documenting the field performance of the unit, active monitoring became very handy during commissioning. When we installed our sensors and began collecting data we immediately noticed that the system was not operating properly. The compressor was running, or the compressor was not running and all of the domestic hot water supply was being delivered by the electric resistance elements. The field installer did not notice the error because the evaporator fan and circulator pump was both operating continually, giving the sense that the unit was operational. The only way that they would have noticed is if they remained nearby for a long period of time to notice that the unit wasn't shutting off. The installers were not familiar with the technology and the control interphase did not give any kind of immediate feedback or warning light to alert to the issue.

Another issue that was noticed was the temperature feedback for the heat pump was low and it was determined that a sensor was improperly installed in the tank. So we had them fix that during the first monitoring phase. Then the manufacturer supplied recommended temperature settings for the operation to make more use of the heat pump water heater, but we noticed during monitoring that the heat pump cycles were short and frequent and we suggested different points in order to optimize and extend full load operation. But during this monitoring period we decided to continue with the manufacture's recommended settings, and that's reflected in the data that's reported.

About eight months into monitoring a bearing in the evaporator fan failed and the unit shut off, defaulting to electric resistance heat. This failure again would not have been noticed for a while, at least not until the electric bills were received, but since we were actively monitoring we reported the failure and it was repaired quickly. This failure is likely unique and not indicative of typical unit reliability, but it's worth noting that there was not at any time any sort of feedback that would allow the building maintenance staff to recognize that the unit had failed. So later on the manufacturer installed a service warning light on the panel and personnel began doing daily inspections. Finally, near the end of monitoring, the controller lost its set points and the unit again defaulted for electric resistance heat and so we were still monitoring and were able to report that error rather quickly.

There are 45 units in this housing complex and the findings of this monitoring were shared with maintenance workers in hopes that the other units would be commissioned and operating the same.

So this slide shows a full year’s data from the end of our commissioning in October through the end of January 2013. Now, student occupancy is atypical to normal multifamily occupancy behavior. Students often have long daily schedules and are absent during observed holidays and breaks. As you can see the demand drops down significantly during the Christmas break and reduces again during the summer. In the tail end of monitoring you can see where the heat pump water heater had shut off due to its lost set points and it took some time for the building maintenance staff to respond because it was during the Christmas break. And when the sessions resumed again in January, other maintenance items took a higher priority.

In the second chart you can see the difference in energy consumption as related to the demand for both electric resistance operation periods and they heat pump water heater operating periods. There is almost 2.5 times the amount of energy use when we defaulted to electric resistance as compared with the heat pump.

The data from the monitoring period was then compiled and filtered for full load operation to determine the relationship of capacity and power to both entering water temperature and outdoor temperature. The relationship was developed using linear regression and the resulting map was formatted in such a way that we could use it with TRNSYS model for simulations. TRNSYS is a transient simulation suite. The map is slightly limited in that the entering water temperature was on average 125 degrees and the flow rate 20 GPM, conditions that are different than the standards ratings.

So the TRNSYS model using the monitoring performance map was developed to ultimately perform a full year’s simulation and then run in various climates. The model is structured very similar to the plumbing structure of the system but different heat pump water heater models were used to simulate both start conditions in which the capacity is degraded, freezing lockout in which the compressor would be shut off -- and in this case it was 32 degrees, and evaporator fan modulation that was observed at high outdoor ambient conditions. The heat pump controls from observations, as well as from literature, were programmed into the model and a simple building structure was used to supply the ambient condition for the storage units, which were located inside of a service room. The model was first validated using weekly sample data from both the winter and summer sessions and then transitioned to a full year simulation using a typical hourly demand profile from the ASHRAE service water heating guidelines. When the annual simulation was performed in various climates, the hot water demand was varied slightly to account for the variation in distribution losses.

Finally, the annual simulations were then run replacing the heat pump water heater with typical electric storage and natural gas storage models. The electric storage model was a multiple node -- that's 8 node model, that used the same ratings from the storage units observed in the field. The natural gas model is a similarly sized unit with the nominal 80% efficiency. The climates compared include hot-dry Sacramento and Phoenix, hot-humid Houston, cold Denver and Chicago, and marine Seattle.

The annual operation costs were compiled using the state average utility pricings for electric and natural gas as reported by the U.S. Energy Information Administration. In all cases the heat pump shows significant advantage over the electric storage units, between 49 and 59%. Now the incremental costs of this installation were about $500 per apartment. Though in this case, as we were told by the developer, by going electric there were additional infrastructure cost savings by not having to supply natural gas and venting as well as freed floor space by not having individual unit heaters. So they told us for this installation they assumed a zero incremental cost.

This chart here shows just the operational cost comparison but if we were to assume a $10,000 incremental cost, relative to a central natural gas system, the annual savings would have to be about $1,000 in order to get a payback of 10 years. That doesn't include utility or state rebates. In climates where natural gas is both readily available and cheap, this could pose a problem. But in many climates where electricity is expensive, more expensive, the unit operation is more favorable.

To give another view, this map shows the ratio of natural gas costs to electricity. In the green zones natural gas is less expensive than electricity, assuming two COP heat pumps as the transition point, and the blue zones show that heat pump water heaters would be more effective. So for instance, while Georgia, Tennessee, Oklahoma are more favorable to heat pump water heaters, Alaska and California are more favorable to natural gas. Alaska definitely, because it's cold. The areas in white did not have natural gas rate data from the EIA.

So in summary, we noticed from monitoring that the system performance was impacted significantly by cycling. A lot of degraded performance from the cycling, especially during the summer when occupancy was low and there was a lot more parasitic losses, but it was also impacted by lower flow rates and the higher inlet temperatures from the recirculation loop. We suggested that either a dead-band be widened or the controls changed, some to improve or extend full load operation, but maybe even going to a two-stage heat pump water heater would improve that performance. Installation and training is critical. These contractors had not had much experience with heat pump water heaters and there was not, at the time, adequate feedback to alert maintenance of unit failures. That was later remedied. Had the unit been in service and not monitored, it would take months for operators to notice the failures in the utility bills.

So the unit was sized properly to meet the load. There was very little electric resistance usage. It fits with some of the normal recommendations for single-family homes. Finally, the savings are most significant in areas where electric or electric costs are higher than natural gas and cheap natural gas can prove a challenge with the higher installation costs of these units. With that, thank you for your time and I'll pass on to the next presentation.

Gail Werren:
Next up are Jordan and Eric.

Jordan Dentz:
OK, thank-you very much. This is Jordan. Let me go to the beginning of the presentation. Alright, thank-you. Let me minimize this. We're going to be discussing a variety of control systems for central domestic hot water systems and a field test of two control strategies in four buildings in New York City. The research was sponsored by the Building America program as well as the New York State Energy Research and Development Authority and lead research team with the Levy Partnership and CDH Energy.

I'm going to provide the introduction and then Eric is going to pick up for the balance of the presentation. Just some background -- energy consumption for water heating in multifamily buildings is a significant share of site energy use and 20% nationally, according to the Energy Information Administration. Energy efficiency of central hot water systems and multi-family buildings can be very wasteful. This diagram shows, for one of our sites, and this is typical for other sites that we've looked at, and others have studied, that perhaps only half of the energy that is generated at the water heater actually reaches the end user as useful energy. The rest is lost to either boiler losses, standby losses, recirculating losses, or other losses.

The present context, the central recirculating domestic hot water systems are common in multifamily buildings. Often the recirculating pumps run 24 hours a day and many times without any sort of controls. That increases energy consumption in two ways. The primary way is that heat recirculating, hot water recirculating, through those lines loses heat to the surrounding environment through radiation. Minor energy loss is also due to the pump energy, which is running continuously.

Why do we have these systems? Primarily to reduce wait time at the tap for users. You can see this diagram, the two diagrams, here shows typical systems with a water heater. There is a supply line and then a return line that brings it back to the water heater. That pump keeps the line hot so that wait time at the tap is minimized.

There are four major types of controls, and I'll just review quickly each one. Timer controls, which Beth mentioned in her presentation, are simple time clocks that turn the recirculating pump on and off according to a time schedule and generally the time schedule should approximate peak periods, when water pressure will provide the flow. When a user demands hot water during an off period, however, they're going to have to wait until the water reaches the end use -- reaches the point of use.

Temperature control is another form of recirculating pump control and it automatically turns the pump on and off based on the temperature set point, usually around 120 degrees, with a sensor on the return line. This will use less pump electricity because it turns the pump off but it won't do much to reduce the radiative losses from the pipes because the pipe temperature will remain at 120, or whatever the set point is.

A third type of control is temperature modulation control. This doesn't control the pump but it controls the supply set point and based on a schedule it increases the supply temperature during peak times and reduces it during off peak time. So energy can be saved by having a lower overall water temperature and reducing the heat loss in the pipes that way, but the pump does run continuously.

Finally, demand control is a newer control strategy and one that we tested in this study. It uses two pieces of information -- real-time user demand based on a flow switch installed in the make-up water line and return water temperature at the pump. And the pump will only run if there is demand and the temperature at the return is below a set point, typically below 100 degrees. Furthermore, the pump will run if it has been off for five continuous hours.

I'm going to turn it over to Eric to discuss our project.

Eric Ansanelli:
Thank-you, Jordan. So we set out with this research project that answers several questions. We wanted to know how the control strategies, temperature modulation and demand control, would compare to constant pumping and to each other in terms of pure energy savings. We also wanted to see what the cost-effectiveness of this would be as a retrofit to a multifamily building and we wanted to also estimate and include the interactive effects between the DHW energy savings and the heating and cooling loads that we talk about in a little bit, and how that might diminish paybacks or not.

Finally, the potential complications we were hoping to identify and resolve, in doing these installations and testing these controls, any kind of problems that occur to make sure they were working well. So previous research, there's a good spread of savings, I'm sorry, energy savings attributed to the controls and they generally range around 10-20% with some higher than that.

Here's a quick profile of our four buildings that we installed the controls in. You can see we had data collected over all four seasons in all the buildings. Three of the four had storage tanks and one had a tankless coil that ran through the space heating boiler and sent hot water to the mixing valve. We'll talk a little about the mixing valves when we get toward the end of the presentation, typical size buildings. And one other thing to consider is these are all, again, in the New York City area. It is a heating dominated climate and that has two effects, both for the space temperature and traction, and also the incoming water makeup temperature varies quite a bit seasonally and there is also some wintertime increase in usage. It's just a behavioral thing among tenants. That increases fuel consumption.

Here's a typical schema from one of the buildings showing where we installed temperature, flow rate, and equipment run time sensors. Actually there is a tiny mistake there. This is a building we had an electronic mixing valve that says mechanical small float.

So the two test strategies and again demand recirculation, temperature modulation. These were run in succession for a period of 1-3 weeks a piece, continuously rotating along with the baseline continuous pump operation over, again, four seasons. And you see there in the description of the temperature modulation controls we had a peak temperature of 125 degrees Fahrenheit and off-peak temperature of 110. I have "peaks" in quotations because you'll see in a few slides that the idea of peaks and off peaks did not come out quite as we had assumed. I'll discuss that a little bit.

These two graphs illustrate the point of varying consumption over the winter and summer season. So you can see in the top left the makeup water temperature coming from our surface water reservoirs in upstate New York varies by quite a bit, about 30 degrees Fahrenheit from summer to winter, and also on the bottom right graph you can see that gallons per day per bedroom roughly increased by 30-40% from summer to winter and that's behavioral.

Can you go back one, actually?

So, in order to count for, you know, when comparing the baseline in the continuous -- across these four seasons and to the test controls, we wanted to make sure we were measuring fuel savings attributed to the controls and not to these effects. So we normalized the boiler run time data for three parameters: the gallons of DHW consumed per day, the incoming cold water temperature from the city mains, and also from the DHW supply water that was being sent out to the building occupants.

So, for some results. First, the installed costs. These were the costs at the buildings where we installed the controls for demand recirculation. It was approximately $3,000. For temperature modulation it ranged from $2,000 to $5,300. If you already have an electronic tempering valve or electronic mixing valve on site, that can reduce cost of temperature modulation obviously. Again, half of this was due to labor and that can vary.

So here we have the supply and return temperatures for all four buildings for all the test cases. This is what's going to drive the kind of savings that you'll achieve, either in reduced temperature at the tap or reduced radiative losses along the recirculation piping. I think there should be two of these for temperature modulation. We set points of 110 and 125 degrees and these are obviously well above that. That brings to bear some commissioning issues we'll talk about in a little bit and obviously there were some savings associated with these two buildings, but not as much as could have been optimized.

So the recirculation run pump time with demand control was reduced substantially and consistently. This is a good thing. The pump control had the pump running 15 minutes a day. That saved several thousand kilowatt hours a year, even with a fractional horsepower pump, and at local energy prices that's several hundred dollars a year.

These are the measured fuel savings at our four test sites. The top row you can see the typical range of baseline fuel consumption for the buildings and the fuel reduction ranged from, across the three strategies from, maybe 6-15% with the exception, again, for those two buildings where it was only about 2%.

Before looking at the financial payback for these fuel reductions, we wanted to account for the space conditioning interactivity. Rather than going with a model where there'd be more assumptions and more room for error, we wanted to do something simple. Knowing that the interactive penalties or bonuses would be some fraction of the DHW energy reduction, we simply applied best and worst case scenarios that gave a reasonable range and conservative range of heating penalties, including bonuses, based on the average typical weather for a region and the typical efficiencies for the HVAC equipment.

So, for the central scenario, about a third of the DHW fuel reduction was considered a potential heat penalty that reduces the total savings. Cooling bonus, again the different units, kilowatt hours versus therms. All the buildings were natural gas, comprised about 10% of the total dollar savings, and maybe somewhat counterintuitively, the reduced pump electric attributes a substantial amount to the dollar savings. It's about a quarter.

Let's take a look at the simple payback. Here you can see for the four buildings annual costs, including the recirc pump electric ranged from about $10,000 to about $30,000 and, now inclusive of the interactive effect, the cost savings on an annual basis averaged about 9% for demand controls, only 3% for the temperature modulation, and 12% for both combined. In the worst case scenario for the interactive effects that we showed on the previous slide, you are still getting less than four years demand payback -- payback for demand control, which is pretty good, and 21 years for temperature modulation, which is not so good especially considering the control is said to have a useful life of maybe 15 years.

Now for the conclusion of the next few slides we'll take a look at some of the take-home lessons that we learned in conducting the study. So, demand profiles, this we found very interesting, very intuitive. This relates back to the utility of temperature modulation control, which relies on the assumption that you’re going to have pronounced peak periods in the morning -- when everybody is getting ready for work, and in the evening when everyone comes home and cooks meals. You don't see that and these are typical graphs and we looked a lot at 24-hour draw profiles across the four buildings. If anything there is a period of low consumption that tends to happen for a few hours in the early morning. This somewhat defied temperature modulation as a strategy. It's possible that a constant, using the constant lower bounds temperature satisfied people using their hot water at all points during the day.

Commissioning -- this is another important take-home and something mentioned briefly in the previous presentation. These three graphs show temperature modulation, typical 48-hour supply temperatures, for the three buildings where we installed these controls. What's happening in the first graph for site A you can see that the lower bound is indeed hitting that 110 and 125 degree set point that was programmed at the control, but for some reason the boiler is responding to a 30, 20-30 degree differential and raising the temperature well above that and, of course, the average temperature was skewed up and the savings were diminished.

For site B, this is the building that had the mixing valve and not a storage tank. Actually, it looks like it did a pretty good job of maintaining the desired temperature set points. 

At site D, again with a storage tank, the lower bound temperature seems to be, at the upper, skewed well above 125 degrees and there is diminished savings. This graph reiterates and illustrates that correlation between boiler run time and make up water temperature, as it varies for summer and winter. You can see a strong inverse relationship here.

So to sort of bring this back to the prior studies that we found in our background literature, we've shown about an average of 9-14% fuel savings from the demand and demand with temperature modulation controls, respectively. Prior studies showing somewhere in that range, plus higher with the manner of circulation, but again they did not account for interactivity and that might be responsible for some of the effect you are seeing here.

Another common concern that we frequently hear is that non-continuous flow that you'd have with the demand control does not work with mechanical mixing valve and this is a popular local brand. It will even void the warranty if you install demand circulation controls, the idea being the non-continuous flow will stress the valve to the point of failure. With electronic mixing valves restrictions aren't as severe but there is still this risk that, after a period of stagnant flow and the flow resumes, there is the potential for hot boiler water to be sent through the valves to the tap at dangerous scalding temperatures. There's ways to address this. There's several manufacturers that produce mixing valves that are approved for non-continuous flow and there's also this idea of a dummy recirculation loop, which we have not tried yet. The idea would be to take a very small pump and install a small recirculation loop that runs just in the boiler room, doesn't incur much of an energy penalty, but allows the valve to see continuous flow.

Speaking to the electronic mixing valve issue of potentially sending out dangerously hot water with non-continuous flow, we did not have fine enough data. We had five minute average temperatures to look at where we are going to find maybe 10 degree, sorry, 10 second in-flow data and get a better picture of this. You can see here that both with continuous flow and with the demand recirculation pump controls. You are getting temperatures that are well above that 125 degree set point. So, that's something we'll be looking at in the future.

Another possible issue that was presented to us is this idea that, with a storage tank and demand recirculation controls installed, you can potentially send very hot water at regular intervals to the supply pipes. So the top graph, these are both from the same building, is both supply temperatures coming from the tank. You see this signature sawtooth pattern and that is characteristic of the tank water being mixed by the recirculation pump that's operating continuously. So you can see it steadily declines as more make up water is added before the boiler fires. In the bottom graph this is what you'd find with either no recirculation system in place at all or with the demand controls running the pump very infrequently. You get more of a flat plateau. This is because the water in the tank is stratifying. The hot water is coming in from the boiler to the top of the tank. It's less dense and staying there at the top of the tank and as the make-up water, which is more dense, comes in at the bottom of the tank, it stays at the bottom. Without the pump to circulate it back and forth, as the 10-inch draw, water comes from the top of the tank. The issue here is that because the Aquastat is placed fairly far down the tank it might sense cooler temperatures with demand control pump and call for the boiler to run and the boiler might send hot water to the top of the tank or it could short-circuit across and go up to the tenants. We've circled two parts on the plot here, which may show this is happening but we're going to be looking more at this in the future and our next project that Jordan is going to talk about shortly and confirm whether this is a prevalent thing or this is just a one-off.

Finally, a word about legionella. The temperatures that favor legionella occur probably in every DHW system at some point on a branch or a dead lake. We're not sure if doing demand or temperature modulation controls and cooling the recirculation, or cooling the tank temperatures periodically, if this poses a risk or it exacerbates that. Now, government agencies are split on the issue with OSHA discouraging demand controls specifically. Other agencies require it. More research is needed specifically to domestic hot water heater systems.

To conclude, an important piece, thermal comfort or occupant comfort - all the controls were implemented without complaint. To be specific, over the four buildings, over the year monitoring period, we did have one complaint to the super that was registered and he did not make any adjustments to the controls and the complaint did not come back. So, we take that as a pretty strong indication of comfort and satisfaction. People will let you know if they are not getting hot water in order to take a shower.

Also, as I said, the draw profiles and demand profiles did not show the pronounced peaks and troughs of on peak and off peak usage. That may indicate that a constant lower temperature would be ideal in place of a temperature modulation strategy. The cost depended on the existing configuration of the building but even then the paybacks for demand control were favorable and again, the scales of building size and also fuel type. These buildings are using natural gas, more expensive fuels, would make the payback more favorable. Jordan?      

Jordan Dentz:
So, I just wanted to mention that we're continuing to work in this area and we're moving forward with a rollout of 40 buildings to install demand controls only. This is sponsored by NYSERDA and we'll be also looking at optimizing supply temperatures and those electronic mixing valve issues that Eric mentioned. So if there are any buildings out there in New York state that are interested in participating in this, please let us know. That is the end of our presentation. Thank-you.

Gail Werren:
OK, thank-you for those outstanding presentations. We have time now for a few questions. We already have some great questions from the audience and you may submit additional questions through the question pane on your screen. Our panelists will answer as many times, as many questions, as time allows.

We have several questions here for Beth. I'll go through those first. The first one is -- in the modeling, was the demand profile based on the measured data for the student village or was it a standardized profile?

Elizabeth Weitzel:
At first for the validation part of the modeling we used profiles from monitored data, but when we went to the annual comparisons we went to a standard draw profile that was informed by the ASHRAE service water heater guidelines.   

Gail Werren:
OK, the next question was -- how many apartments were in the monitored building and how many people lived there?

Elizabeth Weitzel:
Twelve apartments and ...

Gail Werren:
I'm sorry, could you repeat that, Beth?

Elizabeth Weitzel:
Sorry, can you hear me? Twelve apartments and 36 occupants.

Gail Werren:
OK, and then another question is -- is there a recirculation loop that serves each unit or only one central loop?

Elizabeth Weitzel:
One central loop.

Gail Werren:
And then a couple more questions -- what were the steps in the model validation effort?

Elizabeth Weitzel:
The model validation effort -- we started with developing a performance map and then we built the model and ran it with sample, one week, monitoring periods in the spring and the summer, or in the winter and the summer, and we compared that against monitored data and drove it down to less than 1% of variation.

Gail Werren:
And what level of field COPs were observed and how did it vary seasonally?

Elizabeth Weitzel:
We saw monthly average COPs between two and four. The lower happened mostly in the summer where there were higher parasitic, lower loads. Since there were lower loads there was a lot more short cycling and so the performance was degraded.

Gail Werren:
OK, thanks, and now I have several questions for Eric and Jordan. Have you looked at the overall energy impact of the best central system compared to the water heating in individual units?

Jordan Dentz:
That was not part of this, to compare individual unit to central water heating systems was not a part of this study. This was primarily a retrofit based study so that probably would not be within a normal retrofit scope, unless there were significant problems with the existing system. No, we did not look at this in this study.

Gail Werren:
OK. Was there an increase or decrease in tenant satisfaction with demand control?

Eric Ansanelli:
So, like we said towards the end of the presentation, there weren't any formal complaints. We also surveyed the tenants, you know, in addition to periodically checking the superintendents, and what we found was there weren't any complaints attributed to demand control, but some people found temperature modulation ... what would you say ... they did not really like that the temperature fluctuated throughout the course of the day. We did not call it a hard complaint but it seemed to reduce satisfaction somewhat -- the temperature modulation specifically. I don't demand control yielded any dissatisfaction. Of course, there's a question. The fellow who has to wake up at 4 a.m. to be at work very very early, the first person that is going to take a shower when the pump has been off, presumably, for some time -- up to five hours, how long does he or she have to wait for the water to get hot if the pump is pumping something like 20 gallons per minute? We don't have data on that and that might be something that we look at in the next project. But again, no formal complaints.

Gail Werren:
The next question is -- what was the typical flow rate for demand control pumps?

Eric Ansanelli:
So we spot-measured this a few times at the buildings, just at the beginning of the study. That's for the flow rate of the research pump. I think somewhere between 12 and 22 GPM, depending on the building.

Gail Werren:
And then another question -- were domestic distribution hot water pipes insulated in the New York studies on the four buildings?

Jordan Dentz:
Well, two of the buildings the ... well, all of the buildings they looped the piping in the basement, the exposed piping, with insulated. The risers in two of the buildings, buildings A and B were not ...

Eric Ansanelli:
They were older.

Jordan Dentz:
They were older buildings. In buildings C and D, those pipes were well-insulated and it was only a three-story building so the risers were not recirculating. That might explain, partially explain, the lower savings of buildings C and D. Those pipes, the recirculation, was entirely insulated and those buildings.

Gail Werren:
OK, if my building participates in this upcoming study, do we get to keep the controls after the study is complete?

Jordan Dentz:
Yes. Yeah, the study is structured so that the buildings pay half the cost of the installation and labor and we would monitor the performance for a couple months following installation and then use that data for the study and the building would retain the system.

Gail Werren:
And then another question -- what kind of flow sensors did you use and why does per bedroom usage increase so much in winter?

Eric Ansanelli:
Those sensors we used ...

Jordan Dentz:
We used a few different types of flow sensors. So there is the flow switch, which are part of the demand control system and that's a simple switch. It doesn't measure the flow. That's provided by the manufacturer of the innovator -- command control system. Then to measure flow, we used two different types of systems based on the size of the pipes we used and Ultrasonic flow sensor for the larger buildings. For the smaller buildings we used Omega flow sensors. Then Eric, do you want to address the increased usage in the winter?

Eric Ansanelli:
Sure. So we think, and in doing some of the literature, this seems to come down to behavioral usage -- people taking longer showers because it's colder outside. No firm answers. It's a pronounced trend on the graphs. You can see all four of the buildings showing that similar phenomena.

Jordan Dentz:
It does have implications for utility bills disaggregation.

Eric Ansanelli:
Again, heating dominated climate, it would be interesting to see in a warmer climate how that would look.

Gail Werren:
Then another question -- can you repeat the primary reasons why you believe the buildings in your study got less savings from the demand control than past studies reported?

Eric Ansanelli:
The primary reasons probably came down to -- you know, part of this could be the supply temperatures sent to the building. To have 44% savings from demand control I sort of struggle for a reason as to how you could lose so much of that energy from the incoming fuel energy just to the demand recirculation loop. It could be a different configuration in the building with a recirc loop that runs a primary loop in the basement and secondary loops that branch all the way up to the top of a very tall building. I'm not sure but the difference attributed between our 9-12% and the other lower estimates, you know 10-20%, I would guess interactive effects. The other studies did not look at interactive effects with space conditioning.

Jordan Dentz:
It's also, I should point out that all the previous research on demand controls that we researched as part of our literature review were in California so the buildings and climate were quite different than what we worked with here. We have older, larger, or taller buildings but different types of mechanical systems generally.

Gail Werren:
OK, those are all the questions we have for now. Panelists, before we take our quick survey, do you have any additional or closing remarks you'd like to make before we close the webinar?

Jordan Dentz:
No, none here.

Eric Ansanelli:
No.

Gail Werren:
OK. We'd like to ask our audience to answer three short questions about today's webinar. Your feedback will help us to know what we are doing well and where we can improve.

The first question asks whether the webinar content was useful and informative. To answer click on the radio button right in the GoToWebinar panel. The second question asks about the effectiveness of the presenters. And the third question asks whether the webinar met your expectations.

Thank-you for taking our survey. Stay tuned for the next Building America webinar scheduled for February 12 at 3 p.m. Eastern time. The webinar topic will be part one of high-performance enclosure strategies for unvented roof systems and innovative framing. Registration will soon be open on the meetings page of the Building America website. On behalf of the Building America program I'd like to thank our expert panelists for presenting today and our attendees for participating in today's webinar. We've had a terrific audience and we appreciate your time. Please visit the Building America website to download a copy of the slides and learn more about the program. We also invite you to inform your colleagues about Building America resources and services. Have a wonderful week we hope to see you again at future Building America events. This concludes our webinar.