Low Load Heat Pump For Wall Assembly Test Unit
Hey all,
This is a bit of an “out there question” but I have been appreciate of this forum and all the advice and knowledge that is on it.
That being said, I am an Architectural Engineering student continuing on in Grad School this year passionate about wall assemblies and embodied carbon, energy efficient buildings, energy storage, grid connected systems, and everything in between.
We are building a “Unit Test Cell” (4’x8’x7′) that can be used to swap out different wall/roof assemblies and test the performance. Everything from hempcrete to DIS (Dynamic Insulation Systems) or PCM.
The biggest challenge being how to simply control the environment for such a low load (~3kbtu/hr cooling and ~4.8kbtu/hr heating). Just 1/2″ plywood on both sides of 2×4 uninsulated cavity walled 24″ O.C. I feel confident about the loads both with hand calcs and modeling with DOE2 and EnergyPlus; however its still just such a small space at 32 SF.
We don’t want to spend a huge amount of money and I have explored heat pump ERVs like ephoca or CERV2. Just a Hydronic heating and cooling loop but that’s a bit harder to control simply. A window AC and electric resistance and tying the thermostats together. And then the 6000Btu Mistubishi (which I think is probably the best bet since it ramps down to 1700btu).
Does anyone have any suggestions for a simple way to control the temperature for such a low load or any products available?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Whatever your choice it will need to perfectly adjust to the load and achieve stability. I would think for the sake of scientific experiment where all possible variables are eliminated, the control circuit should be something that is thoroughly understood and infinitely adjustable, and the energy flow medium should be easily quantified and accurately measurable.
Given these criteria, my thoughts point towards a hydronic loop and away from a COTS (Commercial Off-The-Shelf) heat pump unit. Controlling and measuring the energy flow of a hydronic circuit is very easily done (and turn-key lab-grade systems may even be available), but I don't believe the same can be said for a bundled heat pump.
Curious on your (and others) thoughts.
Appreciate the advice and I definitely agree towards the hydronic loop but part of the issue is supplying hot and cold water because the unit will be set up to be moved indoors and outdoors frequently and eventually to another portion of campus as apart of the new BEST (Building Energy Smart Technology Center).
The gold standard for measuring heat input is electric resistive heating. You can easily measure volts and amps instantaneously and calculate watts, convert to BTU or joules as needed.
For cooling, I would say hydronics are the way to go, it's straightforward to measure supply temperature, return temperature and flow rate, which gives you BTU's. If you use a buffer tank you can keep supply temperature relatively constant so it's just a matter of varying the flow rate.
In either case, in order to get measurements in a reasonable amount of time you want to keep your conditions as constant as possible, and then vary the flow until the temperature stabilizes. So what you want is a room within a room: an outer room where you keep the temperature constant using traditional climate control, and an inner room that you try to keep at some different temperature, and you measure the heat flow required to keep the inner room at that temperature.
Awesome I appreciate the advice and suggestion. I have been saying we should just do electric resistance and although I haven't set-up a control system before I feel that once you get the ball rolling it's probably not too challenging.
We will be using both the climate control lab test room here at the University as well as outside ambient conditions. The outdoor conditions are a bit more challenging because the desire is to have it be mobile and easily moveable but part of me thinks that's counterintuitive to coming up with a real and accurate test like you're suggesting.
What's the point of doing outdoor ambient measurements? As you said, it makes it much more difficult to make consistent measurements. One of the base principles of the scientific method is you try to eliminate as many variable factors as possible and if possible only measure one at a time. If the thing that you're trying to measure is wall assemblies, you want your experimental apparatus to have everything constant except for the wall assembly.
If you're trying to measure the range of local weather conditions, the National Weather Service has excellent historical records.
If the primary performance parameter you're trying to measure is resistance to heat flow (R-value), the device you need is a hot box apparatus. ASTM c1363 describes the standard test method (in the US) for testing wall assemblies using various types of hot boxes, as well as standards for construction hot boxes. This is an engineering problem, not a problem in basic physics that requires design of entirely new experiments, methods, and equipment. We used liquid nitrogen because the ASTM standard explicitly mentions LN2 as an accepted cooling method.
In all cases, the gist of the hot box method is simple: the wall assembly to be tested separates the apparatus's 'hot side' from its 'cold side' . Once the system reaches equilibrium, both hot and cold sides are maintained at constant temperature. Power input to the hot side is measured, and used to calculate the flow of heat from hot to cold side.
For this calculation to be at all accurate, it's essential that the cold side shows as little variation as possible. Using ambient outside conditions as the cold side makes that difficult, to say the least. It is also essential to minimize flanking losses (that is, heat flow around the assembly being tested, rather than through it) and heat losses through the box walls rather than the test assembly. The most expedient way to do that is to surround the test assembly with material of higher R value than the suspected R of the assembly, and also to have fully characterized the apparatus before beginning test runs.
Prior to designing and building our hot box, we looked at a paper by a Canadian team that had developed a box of similar size with a similar budget. They placed their apparatus in a controlled environmental chamber and relied on the (controllable) ambient chamber temps to hold the cold side (which was uninsulated) at constant temp. Using this method, they report their apparatus fell short of the ASTM standard, both in terms of hot and cold side temp fluctuations and box wall and flanking loss percentages. All this is discussed in the paper at the link below.
Not trying to toot our own horn here (too much); just saying what we found was that careful air sealing (note the plastic film covering the wall section in the pics I attached below) AND plenty of insulation turn out to be the keys to a reliable and accurate smaller-scale hot box, just as they are in building performance. If you look at the chart from a sample test run in our paper, you'll see a point where ambient lab temps fluctuated wildly due to someone opening a big garage door into the space during the middle of winter. Temps inside the box (hot and cold sides) didn't budge, because the box interior was isolated from ambient temps by 6" of EPS, and careful air-sealing.
DC Contrarian is right: you've got to reduce your variables to a minimum—ideally to just the one you're testing—to make this work. Fortunately, ASTM has already figured out how to do that, and published explicit directions. Also note that noise in a small scale apparatus is relatively harder to control than in larger scale apparatus, which is part of the design challenge for boxes of the scale you're proposing, and the ones built by our lab and the Canadian team.
Good luck, both with this and the Solar Decathlon.
Hi Paul,
Is what you're building the same or similar to a hot box apparatus? If so, I might be able to help. A few years ago, I designed and built a small hot box for the university research lab I co-direct. It conforms to minimum ASTM standards, and can accommodate a 1 meter x 1 meter test wall. We achieve cold side temps as low as -20° C.
Hot side uses electric resistance heating, for the reasons DC Contrarian mentions. For cooling, we used liquid nitrogen, pulsed through a thermostatically-controlled valve. Liquid nitrogen gave us control of cold side temps within ASTM limits (+/- 1° C, IIRC), with very low upfront cost compared to mechanical refrigeration, at least for the temps we wanted to hit. Operating costs were obviously much higher, but manageable for our initial test runs.
We built the hot box to determine insulation value of wall assemblies made from a novel masonry unit our lab developed, made from drywall waste.
If you're interested, I'd be glad to post a link to my ResearchGate profile, which has a conference paper we presented that includes construction details, budget, and some test results.
--David
Awesome thank you so much for the suggestion! Everyone is bringing up really great points and I really appreciate it and the advice and glad I asked!
Interesting about using Liquid Nitrogen as well and I will look more into that and hadn't even thought or considered that.
Yes I would appreciate the link!
Here you go, Paul:
https://www.researchgate.net/profile/David-Drake-12
https://labs.wsu.edu/design-ecology/
If you don't need to go quite that cold you could use ice or an ice/salt mixture.
Just keep in mind that a ton of cooling is one ton of ice per 24 hours.
Thanks for that I forgot about that actually and remember wondering about when I first learned about tons of cooling and had to think back to thermo and phase change
I have not been the primary investigator on various climate chamber/hot box projects (grad school and working career), but I have overlapped with those projects enough about them to offer a few suggestions.
First--regarding the heating and cooling systems--simpler is always better, if you can get equipment that matches your required capacity/temperature ranges. So resistance heating and off-the-shelf cooling units probably make the most sense. Do you want to spend your lab time figuring out control strategies, plumbing, and hydronics systems, or actually doing your experiment? Per other comments, resistance heating is incredibly simple to meter. Also, are you controlling humidity on the sides of your climate chamber? That adds another variable.
Second--I don't know how much you've dug into your literature search, but I can only suggest that you want to spend a lot of time on it. It's always worthwhile to know what others have figured out and learning from their lessons, rather than re-inventing the wheel. My colleagues who are now at RDH did a massive guarded hot box test assessing many wall assemblies for heat flow and air leakage impacts--some key reports and presentations here:
Development of a New Hot Box Apparatus to Measure Building Enclosure Thermal Performance
https://www.rdh.com/resource/development-of-a-new-hot-box-apparatus-to-measure-building-enclosure-thermal-performance/
Thermal Metric Summary Report
https://www.buildingscience.com/sites/default/files/project/20150618_thermal_metric_summary_report_-_june_2015_update.pdf
BSC Seventeenth Building Science Symposium | August 5-7, 2013 | Westford, MA
https://www.buildingscience.com/sites/default/files/06_Thermal%20Metric.pdf
06. Thermal Metric: The End is Near
Christopher Schumacher
It would be an utter coup if you could get Christopher Schumacher or Aaron Grin on your thesis committee, but I doubt that would be possible given that they are busy consultants/principals at RDH now.
https://www.rdh.com/location/waterloo/
Good luck!
Hey thanks so much for the advice!
I agree that the goal is not to design an HVAC system but just to simply control the environment and spending my time testing without a doubt which is why I think my professor keeps insisting on something simple rather than my creative designs. Part of why I think just the mitsubishi system that ramps down to ~1700 Btus is probably the best bet and the chances of short cycling are minimized in a sense. However maybe I am thinking about this wrong but if we are just measuring the current to the heat pump wouldn't we need to factor in the COP as compared to just electric resistance assuming a COP of 1? We won't be controlling humidity much to my remorse since that is one of my interests which I worked on for my Senior Capstone Project.
Also thank you for the suggestion to do more Lit Research. I need to definitely work on more since I sort of just jumped into the project to build it and find an HVAC system and the design for much of the rest of it had already been done and wasn't my idea for the project. My focus has been primarily on Renewable Energy systems and Mechanical systems and am apart of the CU Solar Decathlon Team for the 2023 Build Challenge so have been incredibly invested and spent busy working on that home and I can say that I have read a ton of articles from the Building Science Corp. We are incorporating a number of ideas we learned reading into our wall assembly and construction.
Anyways I really appreciate it and you have given me a lot to think about. Part of my concern from the start is we are modeling it without insulation and a rather high air leakage. I have been thinking we need the other walls to be as adiabatic as possible and to try to keep it airtight (while also removable) to really be able to test it. Reading some of those papers I feel more validated that my concerns were accurate and I will bring it up to discuss.
Thanks again for all the help, advice, and resources
Given that measuring heat delivery with resistance heating is so much easier than measuring cooling, I would ask whether you really need to do both. If the goal is to measure heat flow with one side at 100 F and the other at 75 F, wht not heat the inside to 100 F and measure in a 75 F ambient? If you think that something about the wall has different heat flow characteristics in one direction than the other, well, that's probably not the case, but you could build the assembly inside out if you wanted to check.