GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Rheem Gen 5 HPWH – COP 4.0!

lance_p | Posted in Energy Efficiency and Durability on

Rheem now has a Gen V version of their Heat Pump Water Heater available, now called Proterra:

Boasts a COP of up to 4.0 now, and looks like Time-Of-Use scheduling as well.

Are there any other HPWH that compare?  I know the Sanden is regarded very well, but it’s very expensive.  I am, of course, assuming this new Rheem will cost about the same as the current model.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. onslow | | #1

    Same question, different twist.
    I have come to question just how useful HPWHs might be in my local environment. Here in the south west, humidity is generally lower overall than most of the country. So low that "swamp coolers" are a frequent choice to cool homes here. Since swamp coolers do exactly the opposite of what A/C- HPWHs do, I decided to try and quanitfy what level of energy is available in lower humidity parts of the country.

    Initially, I got lost in charts and formulas for long forgotten physics that didn't quite seem related to HPWHs, which are simple air conditioners with one coil stuck in a tank of water. An house A/C reference about condensate and the BTUs required to return air moisture back to water seemed to provide a more relevant start point. I may be very wrong, but here goes.

    The A/C reference stated you need to remove 1060 BTU from a pound of water vapor to return it to liquid water at 52F. This is higher than the 970 BTUs needed to make 1lb of 212F water into steam, and back again, and for some reason this value doesn't mesh with the additional 160F drop from the 212F water in physics experiments. But I have used the 1060 BTU figure in the following thoughts.

    The sensible heat available in air appears to be trivial compared to the latent heat in the humidity. The physics books say 1/4 BTU per F degree per lb for perfectly dry air. Likely not relevant to the real world. The latent heat of the water vapor per pound is substantial, as seen by the 1060 BTU value per lb of water condensed. What is not very substantial seems to be the total weight of water in a fixed volume of air, ie. the air inside a typical home in the southwest, as well as low ACH northern homes with balanced humidity.

    Like A/C units for homes, the air passed through a HPWH will only give up part of its sensible or latent energy, so I am looking at the extracted water weight in a fixed volume of air to help judge available energy. The numbers I am coming up with indicate either, I am completely misunderstanding the physics involved or HPWH COPs will be very poor in low humidity environments.

    Reading the chart correctly (always a chance I am not) then air at 70F and 50% humidity has water weight of 0.0079 lbs. per lb of air or roughly 1/8 of an ounce. A pound of air is 14 cu ft on the psychrometric chart, so to gather up 1 lb of water will take 128 times 1/8 oz. to get to 1 lb. That means 128 x 14 = 1792 cu ft of air holding 1 lb. And the pound of water is available only if you can dessicate the air to perfectly dry. Not very practical.

    An 1800 sf home on slab with 8' ceilings has ~14,400 cu ft of volume available. For experimental purposes, no leaks. This is only 8x the volume holding 1 lb of water, so if one dessicated the whole house and harvested the energy of 8 lbs of water, you only have 8480 BTUs. Enough to raise about 11 gallons of water by 90F. Even throwing in double the recovered energy doesn't quite compensate for turning your home into the Atacama Desert.

    It appears that warm moist air from an unlimited exterior source vented through a HPWH is a much more viable option. Am I correct in thinking that the COP of HPWHs in places like a Florida garage will outperform my location by a huge margin? Where is all the magical heat in a 700 cu ft. room minimum specified for many HPWHs? Is the specific heat value of the entire mass of the house the secret sauce? The resident's body heat?

    While HPWHs seem to be working satisfactorily for many in northern locations, the constant question of where to put them seems disconnected from the physics. The moisture to be found in a laundry would appear to have limited BTU value even if every ounce of water vapor from the wet clothes is captured. Sort of like getting one cookie then back to melba toast. Venting HPWHs within the confines of a low ACH structure poses a subset of problems and questions. Even attempting to duct in outside air in good months may be problematic.

    Searching ducting for HPWHs has yielded little quantifiable information. Rheem offers duct adapters from 5" to 8", but no clear CFM or W.C. info. It is hard to imagine the CFM is very high in such quiet units, nor having much push to them. Given the losses inherent in ducting air from any source, I remain suspicious about ducting HPWHs more than a few feet.

    Does anyone out there have hard numbers that quantify the useful and available BTUs successfully being captured by HPWHs? Do Kill-a-watt meters work on them? I suspect that measuring condensate production will only reveal part of the energy capture, but I am not seeing a good physics description for the source of the purported efficiency gains being claimed by manufacturers.

    Where have I missed the BTU bonanza?

    PS: I think I smell a Sanden in my future

    1. Jon_R | | #4

      Roger, redo your analysis with the knowledge that a HPWH will work fine and with a reasonable COP even when it isn't condensing any moisture from the air.

      Also consider that while the HPWH is running, heat is flowing into the house (either from outside or from a heater).

    2. lance_p | | #6

      Roger, since maximum efficiency is the desired outcome, I'm betting these HPWHs lean towards the sensible side of the sensible/latent split and wouldn't be pulling a ton of moisture out of the air unless your house was pretty humid.

      I quickly searched for testing standards and came up with this:

      It's an older 2011 document with heater models that likely no longer exist, but they do a decent job of generating some COP vs temperature curves. BONUS: they also test with reduced/restricted airflow.

      The test conditions seem to be pretty reasonable (77F @ 40% RH is one point). I don't imagine many people's homes would be too much drier than that.

      As far as the new Rheem goes, I'm not sure how a COP of 4.0 in a HPWH compares to the SEER rating of your air conditioner, but they're both doing the same thing. In a cooling-dominated climate I can see a contained unit like the Rheem being more desirable than the Sanden, from an overall system efficiency point of view.

  2. walta100 | | #2

    I can’t tell you much about this unit but do not get to excited about time of use scheduling unless your time of use costs vary from your utility or you are going to be away for days at a time. My guess is you can expect a HPWH to have a longer recovery time after each use.

    It has been a while since I studied refrigeration but from what I remember the sensible and latent is important to the evaporator side of things for comfort point of view but a BTU is a BTU. Once the heat has been absorbed and the refrigerant boiled into a gas it does not matter to the matter to the compressor or condenser where it came from. Yes when you take heat from dry air the temperature will change more than moist air does but I do not think it will that change the COP unless the room temp changes a lot.

    I do think you will have a different COP at 37° than at 137°


    1. lance_p | | #3

      Walter, agreed. If running in heat pump-only mode the recovery times will be long enough that the unit will be running most of the day to recover from three or four showers. Ideal if you plan to use it as cooling/dehumidifying for the house.

      1. Jon_R | | #5

        I get 2.6 hours of run-time per 15 gallon shower. That leaves lots of time when it isn't cooling or dehumidifying (which isn't much even when running).

        1. lance_p | | #7

          Jon, this is about what I figured, that the unit would be running for roughly 9-10 hours per day. 60 gal x 80F = 40,000 BTU / 4200 BTU/hr = 9.5 hours, or just over 3 tons of nearly free AC (during the cooling season).

          1. Jon_R | | #8

            I'd call it .35 tons for 2.6 to 10 hours and zero tons for the remainder of the day. Which is quite dissimilar to 3 tons.

          2. lance_p | | #11

            Reply to post #8

            My poor/confusing wording. You are correct.

            The HPWH in my scenario would be like running a 1 ton AC for 3.3 hours/day. It's about 0.14 tons of AC total, one ton being equal to 288kBTU. Thanks for pointing that out.

  3. onslow | | #9

    To All,

    Well this still doesn't quite address the question of where the BTUs are being taken from.

    COP according to the very lengthy analysis in the link provided by Lance is a measure of the efficiency of the heat pump's ability to transfer BTUs. The quality of the machine if you will. If you use 400watts of electrical energy to move 1600 watts or 5459 BTUs of energy you get a COP of 4. Still doesn't tell you where the BTUs were transferred from.

    The charts show the COP changes in a straight down hill line based on the temperature of the water being discharged into. This is pretty much like what happens to overall efficiency of home A/C when the outside coil is placed in a location that goes from shade to sun. Sun comes round, the coil struggles more give up its heat, the COP goes down. The report shows COP does also vary according to the air temperature supplied, but the downhill line is still tied to the rising temperature of the tank water. It just gets harder to transfer heat across a narrower delta T.

    The charts describing the effect of humidity show surprisingly little effect on COP and its relation to the tank water temperature. Since the quality of the machine is not changing, this does make sense to me. It is still efficiently transferring whatever BTUs it can find. Still no reference on where the BTUs are being transferred from though.

    The report does note a pretty big difference between manufacturers stated recovery rates and the ones observed. This may mean that the manufacturer used a theoretical derivation based on who knows what. The COP for each machine still looks good, variations in the true capacity of each machine tested may be the issue. There were some notable differences in designs.

    The conclusions pages, 60-61 first notes that while HPWHs will work in all climates, warmer climates will benefit more from the "space cooling/dehumidification effects", which kinda sounds like what I was saying.

    They allow that HPWHs will use less electricity than resistive water heaters by 1/3 to 1/2 which is fine. For people with useful seasonal air to feed through them, HPWH will make good sense. I do suspect that some people may think that a COP of 4 rating will translate into 1/4 of the electrical usage of standard electric water heaters. The report shows that will not be the case.

    I am still thinking that mining the heat out of a home in cold climates, whether by harvesting sensible or latent heat does not make energy sense. Especially in dry climates like mine. I have the additional problem of altitude, which effects fan efficiency. This creates a de-rating effect for mini-splits. No reason it would not apply for HPWH.

    In any case, the heat has to be put into a home by some method and expense. Re-stuffing it into a tank of water however efficiently (the COP) seems subject to the variables I tried to describe. If a unit can reduce the air temperature by 15F in a small room, then that drop will need to be made up somehow. It may be instructive to note that pint of water weighs 1 lb, so condensate in that quantity would represent 1060 BTUs captured. If only a few pints of condensate are to be expected in the course of the HPWH's daily run, then is the net gain of "free" BTUs that significant?

    As I noted before, it checks out in the math that Florida has buckets of latent energy embedded in the humidity along with higher temperatures. A garage location would be just fine and electrical usage compared to resistance might handily be under 1/2 that of the regular heater.

    I did like finding out that the big unit tested needed 500CFM to achieve its best operation. Quite the wind machine. A smaller one ran at 300CFM. Alas, still no data on the W.C. and ducting effects. However, the "plugging the intake test" showed remarkably little impact, so maybe the ducting question is moot.

    I am still wondering about how many and where the BTUs harvested are coming from, as well as how many from condensate value and how many from sensible.

    1. Jon_R | | #10

      Most of the BTUs come from the surrounding air and water vapor (in a ratio that depends on conditions). Some come from heat being generated by the compressor motor.

    2. lance_p | | #12

      Roger, the fact humidity seems to have little effect on the performance of these HPWHs suggests they don't run evaporators very cold and have little latent heat recovery. Minisplits generally work this way to get their super-high efficiencies. When operated in their "dry" modes (dehumidify) they are far less efficient at space cooling.

      If a HPWH has a 4.0 COP, 3 units of heat are being taken from the air passing over the evaporator coil and 1 is being taken from the compressor. The compressor sits upstream of the evaporator coil and is cooled by the incoming air.

      Keep in mind when this report was written in 2011, HPWHs were nowhere near as efficient as they are now. This newest Rheem is almost twice as efficient as these older models.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |