ASHRAE Journal:
ASHRAE Journal presents:
Drew Champlin:
Welcome to this episode of ASHRAE Journal podcast. I'm your host, Drew Champlin, Editor of ASHRAE Journal. Today, we will be speaking with Hwakong Cheng of Taylor Engineers, and Paul Raftery of the University of California Berkeley.
The topic will be reducing emissions from heating hot water systems and commercial buildings.
These guys will talk about a study they performed, and how end users can take the results of that and apply it to their own buildings. These two wrote articles that recently appeared in ASHRAE Journal in the December, 2024 and August, 2024 editions. Let me let my guests introduce themselves up. Paul, why don't you go ahead and go first and then Hwakong?
Paul Raftery:
Hi, Drew. My name is Paul Raftery. I'm a Staff Researcher here at the Center for the Built Environment at UC Berkeley.
Hwakong Cheng:
I'm Hwakong Cheng. I'm a Principal at Taylor Engineers.
Drew Champlin:
Let's get to the conversation. Reducing gas consumption in existing large commercial buildings was the project that you guys have worked on for a while that led to some technical articles, and other awards, and stuff like that, and more research and application. Paul, what inspired this?
Paul Raftery:
The original study that kind of kicked this off was a project to look at overall heating hot water system performance in a large commercial building. We took an existing project that Hwakong was working on and we looked at the boiler efficiency. We looked at the input gas measurements and the output hot water, and then we looked at the distribution piping losses, so how much heat was lost from the piping going around the building. Overall, we found that the actual boiler efficiency was really, really low, in the 30% range. There was also pretty large losses from the hot water loop from the distribution losses as well.
Those were pretty alarming findings for us, but it was just one building. Though with lots of experience, seeing buildings running their heating hot water systems pretty much continuously all year round at high temperatures, and we'd seen some of these issues with short cycling boilers in the field, that was also anecdotal. We really wanted to try and get some funding to dive into this topic in a lot more detail. That was what kicked it off, really.
Hwakong Cheng:
It took us a few years to find finally funding opportunity to explore this and dig in further. A few years back, the California Energy Commission had a program that was looking at cost-effective ways to reduce natural gas consumption in existing buildings, so reducing that gas consumption, rather than just simply fuel switching and putting in heat pumps instead.
This really worked out perfectly as an opportunity to follow up on that original study and explore some of the related topics that we've been sort of thinking about and tinkering with on related efforts. Really, we have to give a big shout-out to the California Energy Commission, the CEC, for really providing tremendous leadership and support for energy efficiency for the last 50 or so years.
Drew Champlin:
Hwakong, what is the significance of this to the industry?
Hwakong Cheng:
Yeah. There are really so many key findings I think that are of interest to industry here, not just reports that you finish and put up on the shelves, but I think things that we can then take, and apply, and share with industry to change how we do things to improve efficiency. One of the key findings was that boiler plants don't actually operate as well as we'd expect in terms of efficiency or that they're capable of achieving.
That's particularly true in older buildings, older boiler plants. We really wanted to understand why not, and what changes in design and operation can we make to improve boiler system efficiency, reduce that gas consumption, reduce energy costs, carbon emissions? That was one key finding.
Another key finding is that a lot of decarbonization efforts today are really focused on electrification, reducing scope one emissions of site emissions at the building, without necessarily considering the impact on the scope two emissions, or the emissions that are associated with the electricity that's being consumed by the heat pumps.
Some electrification projects might not be achieving that net emissions impact that we're expecting, when you factor in how much electricity is consumed and what generators are producing that power.
Paul Raftery:
I think for me, bringing it back to the question, if there's one key takeaway from this project, it's that we were able to demonstrate that there's really large cost-effective savings from in-depth HVAC efficiency measures, that in general as an industry, we're not capturing today.
If you as a client, or as an engineer, or a policymaker are aiming to reduce emissions from existing buildings, those in-depth efficiency measures should be the absolute first priority, because that's where the biggest chunk of the emissions savings really is. That's what we should be focusing our efforts on first today.
Drew Champlin:
Continuing along those lines, Paul, can you tell our listeners about the research approach to this project?
Paul Raftery:
Sure, yeah, I'd be happy to. This was a pretty large research project. It was $1.5 million worth of funding over about a four-year period from the California Energy Commission. We had quite a few different scope items involved there. I'll go over the high-level ones, and then we'll kind of dive into the details later on.
The first was demonstrating what we're talking about here. We went into two large existing office buildings here in the Bay Area, both over a hundred thousand square feet. Over the course of a couple of years, we demonstrated how you can reduce gas consumption in those buildings, and measured the impacts of those changes overall.
The next aspect was looking at the performance of heating hot water systems in the field in a bunch of other buildings. Here, essentially, we were trying to repeat that study that kicked this off that I mentioned earlier. We wanted to look at actual boiler efficiency in some of these older plants in existing buildings, and measuring what sort of pipe distribution losses we were seeing in these buildings.
We also had some full-scale laboratory tests up at Price Industries lab in Winnipeg, where we got to dive into some issues that had been curious about over the years, about how VAV box stratification is affecting some of these systems, and how we can address some of those issues, and improve those designs for new construction or major renovations, where we're really trying to design and operate these systems at much lower water temperatures than traditionally they have been operating at.
Then the last major scope item in this project was trying to gather trend data from building automation systems nationwide. We went into hundreds of buildings, and looked at the automation system, and pulled down heating hot water system trend data, and analyzed that and published that as a data set for others to use as well.
Drew Champlin:
What were the findings at the demonstration buildings?
Paul Raftery:
Yeah, so like I mentioned, we went into these two large commercial office buildings here. They're both in the California Bay Area on the Genentech campus, which was a great partner for us for this project. The existing conditions were that these buildings had a pretty traditional HVAC design, variable air volume, single duct system with hot water reheat provided by boilers, gas boilers.
Each of those buildings had a single, old non-condensing boiler that was coming up to be replaced anyway prior to us kind of kicking off this research project. Just switching out those non-condensing boilers for newer condensing boilers is going to improve efficiency, but we really wanted to go a lot further than that. A lot of the issues that we're going to be talking about here today kind of cascade together and multiply.
If you've got really large distribution losses, and you've a lot of unnecessary loads, and the system is operating too long, and you also have poor boiler efficiency, that all kind of multiplies together. We wanted to see how much gas we could save by correcting all of those issues as much as we possibly could in these existing buildings.
The goal was really to optimize both the design and control of this new heating plant, but also to capture as many of these relatively low-cost controls opportunities at the air handler level, and also down to the individual zone level too.
Hwakong Cheng:
One critical factor found in boiler plant performance is oversizing and the risk of excessive short cycling. We as designers, engineers, we need to size HVAC equipment for a peak load condition, but we don't know what that is when a building hasn't been built yet. We have to do calculations, we have to make assumptions, and be conservative to some degree to make sure that we have enough capacity in case some of those assumptions aren't right.
Equipment also come in discrete sizes. If our calculations end up with a size that's in between, then we're probably going to pick the next size up. For boilers, the end result is often that we end up with equipment that's not sized just perfectly for that peak load, but maybe much as 200% of the peak. In contrast, most of the time, actual heating hot water system, the operating loads, are much, much lower than that.
A typical hot water plant might spend the majority of its time operating at less than 20% of the peak capacity. That really becomes a problem for efficiency when the plant can't operate at such low loads. What happens, the boilers end up short cycling. Each time the boiler cycles off, there's unburned fuel, so the boiler has to go through a purge cycle to exhaust that fuel. That means it's also blowing away perfectly good heat through the flue. Then when the boiler cycles back on, there's another purge cycle, and so on and so forth.
Also, each time the boiler flame cycles off, now we have a hot boiler that's not adding heat to the hot water loop anymore, but it's still losing heat through skin losses. When we have a boiler that's rated at 80% efficiency at full load, the skin losses are part of the other 20%, but those skin losses then are pretty much constant whether the boiler is at full load or at part load. That means that when we're operating at part loads, the skin losses and the inefficiency make up a much bigger fraction of the boiler inefficiency.
This is where all of these issues combined, where excessive short cycling can cause that sort of steady state, 80% rated efficiency that we all assume and expect. That value can really plummet in real operation when we have lots of cycling.
Paul Raftery:
Yeah. We were thinking about all of these issues as we were going to work on these demonstration sites. We advocated for really thinking critically about those peak loads, finding ways to reduce the peak load, which is one of the ASHRAE articles you mentioned earlier there, and then sizing those boilers to really avoid as much over-sizing as was possible, while still meeting the redundancy goals for those particular buildings.
The nice thing about a retrofit is that you don't necessarily have to make as many assumptions as you do in design, because there's a real building. If you have some measured data, you can estimate what those actual peaks are in the real building. Even if you don't have measured hot water load data in the existing building, you might have daily natural gas consumption. You can kind of use that data to infer approximately how oversized the existing boilers are.
We did that in this particular case here. The end result was that we went from a single, oversized and aging non-condensing boiler in each building, to two right-sized condensing boilers. Those condensing boilers also generally offer much better turndown capability than especially the older non-condensing ones. The ones that we ended up selecting were capable of going down to about 10%.
With the two boiler plant, that means we had a 20 to one turndown capability, so that plant can operate for almost all of the loads without short cycling, even down to really low part loads of under one BTU per hour per square foot in this building. When you start looking at those load profiles, which we did in the data analysis part of this project, you see that this is a really common operating point for a lot of these types of buildings in practice.
Hwakong Cheng:
Another key to improving efficiency and condensing boilers is getting that return water, the entering water temperature, as low as possible. To condense, the entering water temperatures generally have to be below 130 degrees, and the lower, the better the efficiency. To help with that, we eliminated unnecessary bypass flow as much as we could. The original primary/secondary flow distribution converted that to primary only variable flow.
We converted a bunch of three-way valves to two-way. We also changed the hot water supply temperature set point control from a fixed 180-degree Fahrenheit set point to one where we reset that set point from 90 to 140 degrees, based on the trim and respond reset logic that's described in ASHRAE Guideline 36. This set point reset strategy means that the plant is only making the water as hot as is needed to satisfy the zone demand, but no hotter than that.
Paul Raftery:
Right. Those measures also help reduce distribution losses like we talked about earlier. You can start to see how these things are cascading together, and there's one knock-on effect from one measure that affects the others. All these measures helped improve the hot water plant performance, but we also looked at low cost control strategies that we could use that would actually reduce the hot water loads themselves.
First, we started up with the air handling units. Those actually had existing trim and respond reset strategies according to Guideline 36, but we found that neither of those resets were really working well due to implementation errors. The end result was that the supplier temperature was mostly being overridden by the operators to fix set points because of that underlying implementation issue.
Also, on the duct static pressure side, the minimum limit was quite high, so it couldn't reset down to very low static pressures. We did a bunch of detailed digging with the folks on site there, and we were able to find and fix the kind of root issues, and get those set points to reset much more effectively in the two buildings.
Hwakong Cheng:
The other thing that we did was really dig into the zone level control. The existing VAV boxes had configurable-only controllers that were 20 or more years old that were doing single-max VAV logic with relatively high zone minimums. Configurable only meaning that couldn't just go in and reprogram them to a new approach according to Guideline 36. We were sort of stuck with the options that were already loaded on those controllers.
I would've looked at that and said, "Well, we can't really do the zone level control in Guideline 36 without ripping those out and putting in new fully programmable controllers," but that would've been outside of the scope of our project and our budget. Again, with more careful digging into the manufacturer's literature, we realized that these old controllers actually had the ability to be configured to do something pretty close to the dual maximum logic that's in Guideline 36, and similar to what's required by current building energy codes.
We did a detailed zone by zone ventilation takeoffs, and set the VAV minimums accordingly, so not reprogramming, just adjusting the set point values in each of these boxes based on the zone calculations. Where the original minimums were roughly about 30% on average or a little bit higher, we were able to drop that by a factor of two, effectively. I just want to note that this doesn't necessarily change the amount of fresh air that's entering the building at the air handler. It's really just about reducing how much unnecessary recirculation there is at each zone level.
Many past studies, field studies, and simulation studies have shown the importance of these low VAV minimums, both for energy impact as well as for improved thermal comfort.
Paul Raftery:
There were a couple other more minor aspects of these field demonstrations, but the end result of that combination of boiler plant retrofits and these relatively low cost controls changes was that we reduced the natural gas consumption in both of these buildings by about 70%.
Hwakong Cheng:
That means there was also a 70% reduction in site carbon emissions.
Paul Raftery:
Yeah. It was remarkably similar in both buildings. It was 69% in one and 71% in the other. It was about as close a replication as you could get, which that doesn't always happen in real buildings. In the original project, we had hoped to kind of sequence the work so we could separately isolate the savings from the work we did on the boiler plant versus all the air side controls, but those demonstrations were interrupted by the pandemic, and then the lingering kind of pandemic effects on building occupancy and how the buildings were being operated.
By the time that the occupancy and operation returned fully to normal, we just didn't have enough time left in the project to be able to isolate the measures. We looked at them in combination. Then I think another key aspect of this is that it wasn't just natural gas savings that we had here. There was also savings to HVAC electricity use because of the air side controls changes. It wasn't that we were switching from using natural gas to electricity, we were reducing both.
Overall, the project reduced utility costs by about $110,000 per year for these two buildings, or roughly about a half a dollar per square foot per year. Then not only that, but we also tracked the hot and cold calls that the operators were getting from the occupants. They had a kind of formal work order system for doing that on site, and we saw a 60% reduction in the number of hot and cold calls after implementing all of these changes. Both the occupants are more comfortable and the building operators are much happier, because they're getting fewer calls from those occupants.
Yeah, we dove into a lot of topics on this, and there is another paper on this that we published in ACEEE, and we will share a link at the end here to all of these resources. In that particular paper, we also looked at the next steps. After you've gone in and reduced emissions as much as possible through efficiency, how would you further reduce them?
The answer is through electrification, using as efficient a system as you can, and in this case, the realistic application would be an air source heat pump. In the process of doing that, it turns out that those in-depth control measures that we implemented actually saved more overall emissions than if we had just outright electrified the existing heating loads without doing any of those controls measures.
Hwakong Cheng:
Sorry, can you say that again in case anyone missed it?
Paul Raftery:
Yeah, so if we had just done those deep energy efficiency measures we did, that actually saved more emissions than if we had just installed heat pumps instead, and for way lower cost. This really highlights that those in-depth efficiency measures are a key step to reducing emissions. Separately from that in the same paper, we also showed that even a relatively small heat pump can capture a lot of the potential savings from electrification.
Even if it's not feasible to fully electrify because you have space, or structural, or electrical constraints, you can get the vast majority of those emission savings from electrification for a relatively small heat pump. For this particular building, we assess that if you size the heat pump to about 20% of the peak heating load, with the remaining loads still served by the existing boilers, you'd have about the same overall emissions reduction as a 100% heat pump design.
There was very little benefit from that remaining heat pump capacity, because the time of day that those loads are occurring, it tends to be much colder outside and the grid tends to be running on natural gas power generators. There's relatively little emissions reduction from that additional heat pump capacity.
Hwakong Cheng:
We are also excited to see the project was awarded first place ASHRAE Technology Awards, both at the Golden Gate chapter and at the Region 10 levels, as well as an honorable mention at the Society level. The last point here is that often, there's a concern with control-related energy measures like these, how long do they last? What's the persistence of those savings?
Maybe they show good savings on day one when we hand things over, but how long do the savings persist, and when do the operators start overriding the control for one reason or another? We were able to go back into these buildings now two years later to go back and check, and we found that the resets were still working and still in automatic control. Particularly the zone level measures, like we applied here are really likely to persist, as those set points are much less frequently changed by operators.
If so, they're done on a single zone by zone basis here and there instead of across the whole building. That was really exciting to see, that it wasn't just while we were watching, that we were achieving the savings, but that so far, they've been much longer lasting.
Drew Champlin:
To let everybody know, this particular honorable mention case study will be featured in an upcoming ASHRAE Journal issue at some point in 2025.
Back to the field measurements, Paul, you guys also did field measurements in several other buildings. What did you guys learn?
Paul Raftery:
Yeah, so separately from those demonstration sites, as I mentioned earlier, we wanted to repeat that kind of research study that kicked this all off, and see if we found similar things in other buildings. There's two aspects of this. One was looking at the distribution losses in the piping, and the other is looking at actual boiler efficiency.
For the distribution losses, in a nutshell, what we did for each building is that we went into the building automation system and closed all of the heating end use valves. We closed the coils at VAV boxes, fan coils, and air handlers, and so on, and then we shut off the air handlers, so there's no air flow in the system. We just operated the heating hot water system as it would normally operate, so maintaining whatever its usual set point and differential pressure that it would normally operate at.
Then we just measured how much heat was required to maintain that set point over steady-state conditions. On average, in the seven buildings, we measured losses of about 0.4 BTUs per hour per square foot. Now, that might not seem like much, but it's about roughly half of typical office plug loads. It's definitely not nothing. I think the key point is that it's continuous. It happens all the time these systems are warm. Because the actual end-use loads in these systems are usually far lower than people expect, this distribution losses can actually add up to a decent chunk of annual heating energy consumption.
When we put it in the context of that larger data set that we collected, it was about 11% of annual heating consumption for the median building in that data set. Many folks will look at this and say, "Those losses are occurring within the building envelope, so maybe it's not really a loss, it's helping to heat the building," but a lot of the time, even most of the time, they're in the return plenum. If you think it through, for the majority of the year, it's not really useful heating unless it's very, very cold outside and the air handler actually requires heating at the air handler. A lot of the time, that heat is just being exhausted out of the building.
Then on the other hand, even worse in the summer, those losses have this double penalty aspect to them. You're both wasting hot water to release heat from the piping into the return plenum, but you're also increasing loads at the chilled water plant, because when it's really warm outside, you're going to have to reject that additional heat from the building. You're paying to pump it in there unnecessarily, and then you're paying to pump the heat out of the building again afterwards.
We published two ASHRAE conference papers on this topic, if folks want to dig into the details. One, kind of looking at the overall seven buildings, and then one where we really dived into a lot of detail. Dave Vernon at UC Davis and his team there really went down into the weeds of looking at one particular branch, and doing very detailed measurements to see if those similar findings occurred at the branch level as they did at the building level. We found similar losses there too.
Hwakong Cheng:
We also went back and sort of replicated that original study that we mentioned at the top. We went back and looked in other buildings to measure boiler efficiency, and found similarly poor performance, an existing oversized non-condensing boiler. Funny enough, during our study, that boiler failed and was replaced while we were still taking measurements. The new boiler was smaller, overall, much more efficient, but even then, it still operated far below that nominal efficiencies that we'd expect because of how much it was cycling below that minimum turndown limit.
We're talking about operating efficiencies of 38 to 60% rather than the 80 to 90% that we generally assume or that boilers are rated at. Again, this is for non-condensing boilers. It seems that the main issues are that boilers are excessively oversized, they have limited turndown capability. The third point is that most buildings spend most of the time operating at really low part loads. The end result of all of these three factors is that these boilers are spending a ton of time, maybe most of the time short cycling. That's where efficiency really drops to these low values that we're measuring here.
Obviously, that's not to say that every boiler plant runs like this, but the handful that we've looked at, they do. There are some other white papers and studies showing similar results. I think it's a really important finding. When your boiler is at end of life, rather than replacing like for like, you might be able to actually right-size the replacement based on real measured loads from that real building, get to better turndown capability, possibly increase the plant performance by a factor of two.
For the same cost or maybe even less, we might be saving energy, reducing carbon emissions, and using a known solution to achieve that.
Drew Champlin:
Paul, what did you guys learn from gathering data from hundreds of heating hot water systems nationwide?
Paul Raftery:
Yeah, so this was a really large effort. Collectively, we reached out to hundreds of contacts and ended up screening literally thousands of buildings in this process. At the end of that process, we were able to gather trend data from the building automation system from 259 buildings, and that's nationwide, though there is a strong kind of skew towards California in there too.
The way I like to think about it is that we now have over 700 years' worth of data from heating hot water systems nationwide, and it's a public data set, so anyone can download it and look through it, as well as just read the papers. With the data set this big, there's lots of interesting things you can look at. We won't have time to go into all of them today, but as I said, the paper's there if you want to learn a lot more, if you actually go to the podcast episode page, I think there's going to be a project website where it links to all of these papers and reports, so you can dive into the details.
Back to the actual study itself, some of the key findings were that hot water plants are operating for much longer than they need to, and much more frequently than they need to. To give you some context on this, for the typical building that we looked at, the hot water plant was operating for over 80% of the hours of the year, and that tons of the buildings were essentially operating their hot water plants continuously.
A lot of these are building types that might require that, like a lab or a hospital, but there was also a large number of more traditional, typical occupancies, like offices, where they're running the plant 24/7, 365 days a year, and they're running in the summer, in the nighttime. We have the data to show that this really is a pretty common issue in existing buildings. We also, because we had the data in the automation system, wherever possible, we gathered BTU measurements on the hot water load side.
We're able to say what the load on that system was, and we gathered those together and published the load distributions. Those results show pretty clearly what we're talking about earlier, that these hot water loads are typically much lower than people expect, and this is in buildings from across the country. These hot water systems are spending a huge amount of the time operating at very low loads. That has a whole bunch of knock-on effects.
It means that you really have to size plants for very low turn down if you want to avoid those short cycling issues that we mentioned earlier, and make sure the plant can efficiently operate at this really common operating point. It also means that there's a really large potential for electrifying most of the annual heating load with relatively small heat pumps.
If, for whatever reason, it isn't feasible to fully electrify with a heat pump at a particular site, partial electrification solutions can actually get a huge portion of that annual heating load, because the loads are so low a lot of the time. We also had for a subset of that, the buildings that we looked at, we also had the actually installed equipment, so we were able to say what was the capacity that was installed to meet those loads were measured.
We have some pretty clear evidence that these systems really are commonly oversized by a really wide margin. When we're replacing them, whether it's you're replacing with an existing gas boiler with a new gas boiler, or you're doing full or partial electrification, you really want to size them for the actual loads. That typically means a lot smaller than I think a lot of folks would expect.
Hwakong Cheng:
We also found that a lot of buildings don't have working hot water supply temperature resets. Even buildings where we know there is reset logic implemented, it's just not resetting for one reason or another. There's potentially an easy savings opportunity there. Related to that, we talked about poor efficiencies in non-condensing boilers, but even in the condensing boilers that we looked at, many of those never actually operated at condensing conditions.
The return water temperatures were too high. Even if we couldn't measure the efficiency, we know, simply looking at the return water temperatures, that the efficiencies weren't what we were expecting. There's a really big opportunity there for us to look more closely as designers, and make sure that in design and in operation, that we're working harder to make sure that these condensing boilers are really living up to our expectations.
Drew Champlin:
Given all this data, you've seen that these issues of excess operation and high temperatures are commonplace in the systems. Paul, what do you think is driving that?
Paul Raftery:
Yeah, so I think operators are commonly running these systems continuously at these high temps because there's a perceived risk that the pipes will leak when they cool down either fully or just a little bit. To be clear, in some cases, this is a legitimate concern. Because of a variety of issues that we won't go into here, there can be leaks when some of these systems cool off, but in many of these cases, the solution is really conservative.
It could be resolved by replacing a relatively small number of components, or maybe the piping connections really did leak a lot throughout the building when they cooled down at your last building, and now you're just running the current building at the same 180 degree Fahrenheit, 24/7 condition just in case, even though there has never been an observed leak here, or maybe you could drop that down to 140 and it would never leak, but that was never actually tested.
I think that's a common driver of this. At our demonstration site, just after we put in the controls to disable the boiler plant at night, someone noticed a slow water leak in the ceiling below the mechanical room. I think the gut reaction was that was due to the plant shutting off, and pipes leaking as they contracted when they cooled down. They started running the plant 24/7. In a different situation, that plant could have easily run 24/7 for the next 30 years or so.
Here, the team had the bandwidth and energy to take a closer look and resolve that underlying issue. It turns out it was just a leaking flange at one of the new isolation valves, and it had nothing to do with the water temperature. Contractors fixed it, and the controls to knock the plant off at night were re-enabled, and everything's working great now.
Hwakong Cheng:
On the other hand, we have a current project where we're updating the control of a central plant, including the boiler plant, and there, they do know that the pipes leak when the loops cool down. In order to do our resets, they did testing to confirm that they could safely reset the hot water supply temperature down to 140 degrees without the pipes leaking. What we didn't think about though was that the added lead lag staging logic for the hot water pumps could cause problems.
It turns out that when the lag pump is off, the short segment of pipe connected to that pump, they cool down just enough that they do actually leak. Eventually, that owner is going to have to address that issue more directly. In the meantime, we run both pumps all the time. There are places where this is a real issue, where cooling down the pipes and the connections does cause leaks.
Paul Raftery:
Yeah. I think it's hard to predict in advance which one of those it is, and what the both short and long-term solution is. I think there's also a lack of understanding of the potential benefits from shutting that system down or operating at much lower water temperatures. Up until we published this study on distribution losses, there wasn't a publication that actually went out and measured the distribution losses in these piping systems.
People kind of idealized it a way that there's no distribution losses, but there clearly are, even in a perfectly insulated system. Now, we have some measurements to show that. Clearly, if you shut that system off, you're going to recoup those energy benefits, and you probably will reduce the amount of short cycling boiler time you have as well. There is a large potential savings here.
Up until now, maybe folks didn't generally understand that so fully, so there wasn't so much of an incentive to investigate and kind of fully fix the underlying issue that was driving those leaks so that you could shut these systems off.
Drew Champlin:
Yeah. Your August, 2024 ASHRAE Journal article was on re-optimizing optimal start and morning warm up. Hwakong, would you like to maybe shortly summarize that for the listeners?
Hwakong Cheng:
Yeah. Most buildings are only occupied during the day. At night, we let the temperatures float down in the winter to save energy. Pretty simple. For most buildings, that means that a lot of the heating is actually done in the early morning when it's generally coldest outside, and we're trying to recover those base temperatures back from the setback.
Because this preoccupancy recovery period is often when heating loads peak and when the most heating energy is consumed, morning warm up really deserves, I think, specific attention to make sure that we're doing this as efficiently as we can. That's, I think, especially true for electrification projects, where the size of the heat pump drives a lot of the cost and the feasibility of the project, and also there's big potential impacts on system efficiency.
Drew Champlin:
What are some of the downfalls of conventional morning warmup practices?
Hwakong Cheng:
Optimal start is an energy efficiency strategy that's been around for decades. We set back the temperatures overnight to save energy by reducing envelope losses, and then we use optimal start logic, basically wait as long as we can to heat the building back up, and then try to recover to comfortable temperatures as fast as possible.
Decades ago, before we had variable speed drives, and modulating capacity boilers, good windows, good building envelopes, that was really the state of the art.
Paul Raftery:
Now, with modern equipment and envelopes, it might be more efficient to recover slowly, right?
Hwakong Cheng:
Yeah. You might not need as many boilers. You might not need the hottest hot water. The lower temperatures might mean better efficiencies for your boiler or your heat pump, reduced pipe distribution losses. This all ties back into the work that we've talked about earlier in this podcast.
Fans might also run at slower speeds, but with cube laws, savings there, there's a certain amount of heat that we have to get back into the building during morning warmup. If we do that as quickly as possible, which is the specific goal of optimal start, that means we're making that peak heating load as large as possible.
Paul Raftery:
That might not be such a good thing nowadays, especially when it's driving equipment sizing for relatively expensive capacity, like air source heat pumps, for example.
Hwakong Cheng:
Yeah. For boilers, there really hasn't been much pressure for right-sizing boilers with conventional practice. The next size up boiler doesn't cost that much more, doesn't take up that much more room. There's pressure on designers to not undersize equipment, but now we know that there are potentially negative energy impacts with oversized boilers. Maybe we can size heating equipment to be a little smaller. We talked about how important that is for boilers, but it's even much more important for heat pumps.
Heat pumps can be five to 10 times more expensive to install. They take up a lot more space. They're heavy, so there's potentially structural impacts from the weight, and maybe electrical impacts too, particularly in retrofits. There's a lot more pressure to right size heat pumps specifically in retrofits. It might even be a make or break issue for decarbonization retrofits.
As we electrify more buildings, the impact of morning warmup on the utility grid is also going to be a big issue in some places, and that's where looking at how we do this warmup and minimizing this peak heating load may have a big impact in the future.
Drew Champlin:
Hwakong, what is an improved strategy, and what did the field test show?
Hwakong Cheng:
Yeah, so we had an opportunity to do a quick field study, really just kind of a pilot test to explore some of these ideas. In a community college building in northern California, we were able to play around with some of the warmup strategies.
On some mornings, we let the building warm up according to its existing control, and then other mornings, we warmed the building up much more slowly. When we compared the two approaches for days with similar weather, we saw a huge decrease in the peak heating load for the days where we did this slow warmup.
Paul Raftery:
Right. Then on other days, we also kept those space temperature set points constant at the occupied heating set points. There were kept at 70 Fahrenheit all night, instead of letting them set back then having to recover. Obviously, there's a bunch more heating energy required, because you have a warmer building and higher envelope losses, but you really eliminate the recovery part of that heating load altogether, and that's what the measured data showed.
The heating peak loads are way lower on those mornings. Of course, you might not want to do that every night because of the energy impact, but maybe you'd consider doing that on the coldest days of the year. For example, you don't have enough heating capacity to recover from setback, or maybe your building is all electric, and the warmup heating load might set the utility peak demand or a situation like that, or it might mean that you don't need to install as much heat pump capacity.
Alternatively, if you've got a partial electric system, you could electrify more of your annual heating consumption with that heat pump on that small hybrid system, if you kind of smooth out those peak loads to the point that they can be served by the smaller equipment.
Hwakong Cheng:
Yeah, again, all of this was just a pilot study. It was just one building, and one climate, and for a limited number of days. The goal was really just to show that some of these effects that we've theorized may be real, and that they warrant further study.
Paul Raftery:
Fortunately, ASHRAE is about to do just that. There's an upcoming research project, RP 1964, that is going to do simulation studies and field tests to evaluate these sorts of factors in much more detail and across a range of different climate zones. That will also be providing recommendations for how you do these types of control strategies that will hopefully go into a future version of ASHRAE Guideline 36. Stay tuned and keep an eye out for more research on this topic in the near future.
Drew Champlin:
You guys, you also wrote an article that was published in the December, 2024 ASHRAE Journal on Rethinking VAV Hot Water Terminal Unit Design. Paul, can you just summarize that?
Paul Raftery:
Yeah, sure. It was a busy few years. This particular element started out with some testing that we were doing in VAV reheat boxes as part of testing out a new duct airflow sensor. We were working on a separate project. That, along with a bunch of field measurements and experience we'd seen in these systems, led us to kind of dig into this more deeply.
We did some full-scale laboratory tests with the manufacturer, and the issue is basically that when you're heating with a VAV box, you control the airflow and you control the valve position to maintain that discharge air temperature at a reasonable temperature, a little bit warmer than room, but not too high, so maybe 20 or 25 Fahrenheit warmer than the room.
Hwakong Cheng:
The original concern was that the air temperature leaving the reheat coils isn't uniform, in that the average temperature might be much higher or much lower, depending on where the point temperature sensor is located or perhaps under slightly different operating conditions.
Paul Raftery:
We wanted to first of all understand whether there is significant airflow and temperature stratification on the outlet of the VAV box. If there was, what was the impact of that? What are the causes, and what could we do about it in future design?
Drew Champlin:
Yeah. When we talk about VAV, we're obviously talking variable air volume, just to clarify that for anybody who may have a question.
What kinds of operational issues exist with typical VAV reheat terminal units?
Paul Raftery:
The testing showed what we actually expected, and that, yeah, there really is significant temperature stratification in the air leaving the reheat coils, and it becomes more so the closer those measurements are to the coils. If we look at a set of temperatures through a cross section of the duct that's downstream of the coil, there were sometimes differences of more than 30 or even 35 Fahrenheit from one point to another.
Hwakong Cheng:
That's huge.
Paul Raftery:
Yeah, and it's too much to try and explain in a podcast, but the short version is that dampers in the VAV box calls a lot of stratification and airflow as they are partially closed to maintain a particular airflow through the box. This velocity stratification can lead to pretty extreme temperature stratification too. That damper position has a big impact.
So, of course, does the location of the point temperature sensor that you're using to measure the flow downstream of the coil and damper. Interestingly, this stratification also reduces the heating capacity of the coil quite a bit, which we weren't really expecting. It was around 10 to 15% of a reduction for two-row coils with a moderately closed damper, which is where most of these dampers would be when the whole air handling unit and system is operating in heating mode in the winter.
One of the interesting outcomes of this is that it turns out that having a good duct static pressure reset is actually pretty important in winter months too, and not just to minimize fan energy consumption. That's because the pressure reset will keep pressures in the ducts low, and that will keep dampers more open to maintain the same airflow through the VAV boxes. That will reduce stratification, and give as much heating capacity for a given water temperature as possible.
You can check out the papers from lots more info in this, including some visuals that explain this much better than words.
Hwakong Cheng:
Another thing that we looked at was how the coil circuiting might be a factor, so how the hot water flows through the tubes and the coil, and how those tubes are arranged. VAV boxes are kind of a commodity item that have been around forever. Honestly, we don't pay too much attention to details like the coil circuiting, but looking more closely at this for a two-row hot water coil, we found that the tube passes go back and forth between the front row and the back row.
Picture is worth a thousand words, but basically, the coil is a heat exchanger that's a mix of counter flow and parallel flow. If we go back to our college engineering classes, we know that we're not going to get as good heat exchange with parallel flow heat exchangers. We created some custom coils to test along with the standard or the stock coils. One of these customs had a single circuit, which means one pipe in, one pipe out across both of the rows, and it was circuited so that everything was counter flow.
Everything else about the coil was the same as the stock version. Our testing showed that not only did we get less temperature stratification from this custom coil, but we also got significantly improved heat transfer. For the same water and air flows, the custom coil had much higher heating capacity, and importantly, better hot water Delta T. For a coil that costs the same as a standard, that's really important for low temperature applications, like condensing boilers and air to water heat pump projects.
I'm also really happy to see that at least one manufacturer has embraced these results and just recently released a new line of VAV reheat coils based on this work in this lab testing.
Drew Champlin:
Hwakong, what are some of the overall takeaways from this project?
Hwakong Cheng:
One main thing is that there are lots of opportunities to implement low cost measures in a large number of buildings. Some of them are completely overlooked, some of them are things that we're doing already, just not very well, and not achieving our intended goals. Maybe we put in a condensing boiler and then we turn the building over to the owner, but never realize that the boiler is never actually condensing.
Why not? What can we change to make that work better? We often overlook opportunities to reduce the loads served by these systems, for example, by correcting VAV box minimum air flows. We can do better, but we have to pay closer attention and know where some of these key factors are.
Paul Raftery:
Right, and implementing those in-depth efficiency measures is also the key to reducing emissions from these buildings. A lot of decarbonization projects lead with and focus on electrification, which is crucially important, but they then overlook the opportunities to improving efficiency. We're going to use heat pumps to fully decarbonize, but sometimes that can be really expensive and add a lot of complexity with space, or structural, or electrical impacts in retrofit scenarios.
Based on the marginal emissions rates from the grid electricity, both now and over time, heat pumps don't 100% reduce emissions if you focus just on the site emissions, they do, but there are still emissions from electricity on the grid, so it's not like a binary thing. In contrast, these deep efficiency measures are reducing natural gas consumption and also reducing electricity consumption. Overall, they provide some really large emissions savings, and often very cost-effectively.
That should always be the first protocol, really driving deep and getting as much efficiency as you possibly can out of the building. Of course, that's not a separate thing from electrification. Ideally, you stage this so that you do those in-depth efficiency measures first, and then you electrify the remaining loads with a heat pump. That will overall yield the lowest cost and overall the largest emissions reductions from these projects.
Drew Champlin:
Well, to wrap this up, Paul, what kind of resources can you guys share with people looking to repeat this in their own buildings?
Paul Raftery:
Like we mentioned, this is a California Energy Commission funded research project, so all of these results are publicly available. We had something like 11 peer-reviewed papers and reports that came out of this study, some with ASHRAE in the Journal and conference papers, other with energy and buildings or ACEEE. All of this work has real, practical implications for how we design and operate hot water systems and HVAC systems in these types of buildings.
If you go to the ASHRAE Journal podcast page for this episode, there is a link to our project website on the CBE, website where all of those papers and reports are linked to, and those are all freely available. That also includes some of the video recordings and some past presentations we've given on this. There's a policy guide and there's a full project report.
For those of you who like digging into the actual data and want to look at things like load profiles and temperatures in real buildings, we actually published the data set from that trend data gathering exercise I mentioned earlier. There's a public data set with data from over 200 buildings that is available for folks to dig into if they so wish.
Hwakong Cheng:
One of those reports was a hot water heating design and retrofit guide. That resource really pulls out a lot of the key findings from the different facets of this project, and distills them and summarizes those findings into a compact guide. It then provides practical and actionable recommendations for what we can do differently to try to reduce hot water loads and improve hot water system efficiency.
Drew Champlin:
Well, that will wrap up this ASHRAE Journal podcast episode. Hwakong Cheng and Paul Raftery, really appreciate your time and your expertise, talking about your project, and how it can affect other commercial building owners as it pertains to heating hot water systems.
Hwakong Cheng:
Thank you. Thank you very much.
Paul Raftery:
Thanks for having us.
Drew Champlin:
I'm Drew Champlin, ASHRAE Journal Editor. Thank you so much for listening.
ASHRAE Journal:
The ASHRAE Journal Podcast team is editor, Drew Champlin; managing editor, Kelly Barraza; producer and assistant editor, Allison Hambrick, assistant editor, Mary Sims; associate editor, Tani Palefski; and technical editor, Rebecca Matyasovski. Copyright ASHRAE. The views expressed in this podcast are those of individuals only, and not of ASHRAE, its sponsors or advertisers. Please refer to ASHRAE.org/podcast for the full disclaimer.