DC Inverter Technology

Discussion in 'General Discussions' started by Corley, Sep 10, 2010.

  1. Corley

    Corley New Member

    I have a Geothermally heated house, with 4 zones. I have a ground loop using a nearby year round stream. So far I've heated the house using the domestic water heater, with great success. Time for a heat pump though, so after searching what is available, I decided to try a new aproach.

    I purchased a 2 ton mini split system, with a DC inverter technology Panasonic compressor. This is an air to air system, I needed a water to water system. It turned out this was the cheapest way to get all the parts I needed. I swapped out the two air exchange units for 50 plate heat exchangers. The box got very small. I'm just in the process of checking it all out at this time.

    Here was my thinking. With a normal 2 ton unit: With 4 zones in the house, when all 4 zone call for heat, the 2 tons would be a good match, providing the right temps into the floor. When only one zone calls for heat, 2 tons is too much, and provides too high a temp into the floor.

    With the variable output unit: When 4 zones call for heat, the unit can run at full output, and supply all zones with the proper temp water. When only one zone calls for heat, the unit can operate at a lower speed, and again, provide the proper temp water into the floor.

    Shoot me down... As I said, this is almost completed at this time, so before long I will know how it all works out. However, I'd like your thoughts on the concept, and execution.
  2. urthbuoy

    urthbuoy Well-Known Member Industry Professional Forum Leader

    Buffer tank?

    A buffer tank generally sorts out these demand imbalances. Are you using one?
  3. Corley

    Corley New Member

    Buffer tank

    As far as a buffer tank, I do not intend to use one for the simple reason that they don't really do much good. Here's why. Say I have a 100 gallon buffer tank. If I'm pumping out of it at the rate of say, 5GPM, (one zone active) it's only good for 20 minutes, and then it's entire contents have been recycled. Oh, you say, but the heat pump is inputting into it at the rate of 20GPM. Well then, the heat pump will have to cycle on and off fairly rapidly in order to not over heat the tank. If I'm pumping 4 zones at 5gpm each out of it, then it only lasts for 5 minutes, unless the heat pump is running solid.

    In one case the heat pump cycles too often, in the other it simply matches the in and out, effectively doing nothing but allowing another place for heat to escape. So, what's the point of the buffer tank really? If you really think about it, the buffer tank has very limited value. I know of one case (Our Cool House web site) where the buffer tank was removed, and the overall efficiency increased a lot. I've decided not to use one, but instead tackle the problem using a variable rate heat pump.

    What I'm really trying to address is the feasabilty of using these 50 plate heat exchangers in this application, or if I'll have to switch to some other type of heat exchanger. Thanks for your input...
  4. Corley

    Corley New Member

    No buffer tank because they are not actually an ideal solution to the imbalance on the system. First, if you consider the size of the buffer tank vs the inputs and outputs, it only provides a very small bufferring ability. Say you have a 100 gallon tank. At 5gpm per zone, with all 4 zones running, 20 minutes and all water has been recycled. So, shortly after the zones start running, the heat pump will commence operation and run almost continuously (assuming it is sized to the max). Nothing was gained at all. Now take the case where only one zone runs, the heat pump will cycle on and off fairly quickly, not a good thing at all.

    I am attempting to create a system that works into whatever load is needed, by varying the compressor speed. These systems (in mini split form) are actually very efficient, even at slower speeds. As far as buffer tanks go, I know of one case (Our Cool House web site) where eliminating it increased efficiency significantly. That's my plan anyway, do a really variable speed pump, with no buffer tank at all.
  5. jrh

    jrh Member

    I would give you an E for effort. If you had all these materials free (a2a heat pump, 50 heat exchangers) and a lot of free time, I would support it.But I think you will have a hard time beating the efficiency and reliability of of even a single stage heat pump with a buffer tank and proper controls.

    How will you be controlling the temperature of the water you are delivering to your radiant floor?
  6. urthbuoy

    urthbuoy Well-Known Member Industry Professional Forum Leader


    I agree that using an inverter controlled compressor is where things are headed, but...

    load matching, in what you are trying to receive is going to be tough with temperatures being returned to a refrigeration cycle. The buffer tank is designed to eliminate the short cycling you mentioned, ensure constant flows, and store heat (not waste it).

    As I understand it, your 50 plate heat exchangers might as well be an indirect tank system.
  7. Corley

    Corley New Member

    [/QUOTE]As I understand it, your 50 plate heat exchangers might as well be an indirect tank system.[/QUOTE]

    Sorry, I don't understand what this sentence means. As for how the temp is controlled, think of it as a mini split unit, which it actually was, and I'm using those electronics. The same sensors as were used to sense air temps are used in this system, and since the air side of things is replaced by a water exchanger, the sensors are now connected to sense water temps. Therefore, when I set the minisplits temp to, say 100 degrees, it outputs 100 degree water instead of 100 degree air. This allows the system to work into whatever load is present, just by keeping the output water temp at the set point. When zone water stops circulating, the sensor indicates water temp above the set point, and the compressor slows down, and eventually stops because we are not removing the heat any longer, and the output heat exchangers becomes warmer than the sensor point. (At 5 degrees above the set point, the compressor shuts down totally.) If multiple floor loops are running, the output water sensor tells the compressor to speed up, and make more hot water. If only one zone circ is running, less cold water is being returned to the output heat exchanger, so the output heat rises, slowing the compressor. It should automativally load match,

    Therefore, it all becomes automatic operation for the compressor, without even any control lines back to the zone circ controls. Just set the output water temp desired, and it will keep that temp in the output exchanger/water line, regardless of how many circs are running. The zone thermostats just turn on and off the circulator pumps, that's it. Almost...

    I have a back door fix for turning off the heat pump if the compressor shut down is not fast enough or effective enough. The board controlling my circs has an output that is active if any zone circ is running. I can use that to also control power to the heat pump if needed, shutting it down completely when no zone circs are running. I'd prefer to let it work automatically if I can, but this still needs testing to determine if I get any overheat situation in the output exchanger after the circs are shut off..

    By the way, the input heat exchanger (connected to the geo closed loop) has freeze protection sensors as well, but my thought is that should never be needed, if I am circulating the proper amount of water/antifreeze mix in that loop. I'm using the mini split's line that originally ran the outdoor fan to control running the geo loop circulator, and it is sized for max heat operation.
  8. urthbuoy

    urthbuoy Well-Known Member Industry Professional Forum Leader

  9. Corley

    Corley New Member

    Indirect fired water heater...

    Well, the indirect fired water heater that you refer to is a way to transfer heat from one liquid heat source to another liquid, but other than that, it is in no way what I am doing. It also provides a storage / buffer tank capability.

    My heat exchangers exchange heat from the ground loop to freon on the input side, and from freon to heat in water for direct input to the zone circs on the output side, and there is no storage / buffer tank involved, as in the indirect fired water heater. Really not the same at all.

    My plate heat exchangers really just replace the shell and coil exchangers of a typical water to water heat pump, and my variable speed compressor (with electronic controls) really just replaces the normal fixed speed unit. The thught here is that this really eliminates any need for any storage / buffer tank.

    At this point, I'm just not sure how well the plate exchangers will work with this setup, since all the commercial units seem to use the shell and coil exchangers. (I got a really exceptional deal on the two plate exchangers or I would have not used them!) I hope that is a clear explaination...
  10. urthbuoy

    urthbuoy Well-Known Member Industry Professional Forum Leader


    To understand the layout.

    The onus will be on the refrigeration system to deal with wider demands this system will require. I won't pretend this is my expertise. A refrigeration forum may have better direction.

    Europe, by the way, already has these inverter controlled water-water units.
  11. Corley

    Corley New Member

    DC Inverter Heat pump

    Believe me, if I could have found a DC Inverter heat pump that was water to water here in the states, I'd have just bought it instead of building it.:)
  12. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    You suggest you are heating the home with a water heater and a geo system.....is this a geo water heater?
    How did you arrive at the conclusion that 2 tons satisfies your heat load? What is the heat load of the house? What was BTU's of geo/water heater currently in use?
    What is your plan for auxiliary if 2 ton falls short in mid winter? Closed loop 2 ton may not be much more than 20MBH when you need it most. I don't know how big your 4 zones are, but..........

    As far as my thoughts on the concept; if you have truely butchered an expensive appliance with no knowledge of load and no calculated net delivery and no auxiliary mentioned......leaves much to be desired.
    Good Luck,
  13. Corley

    Corley New Member

    DC inverter...

    I'm sorry you were not able to read / understand what I posted. [​IMG] All calculations HAVE BEEN DONE, heat loss, ground loop capabilies, loop sizings, etc. As I stated, we spent the last year heating with a dom hot water heater, and fully understand the heat demands, BTUs required, etc. We are now going to reduce energy costs by changing the DOM heater to a DC inverter heat pump. The DOM water heater will continue to be a backup after this conversion. I'm baffled that you didn't get that from my earlier posts, but...
    YES, I have basturdized a mini split unit, by replacing the heat exchangers on both sides of it with water to freon units. Why? Because that was the cheapest way to obtain the parts needed (DC inverter compressor and logic to control it). Will it work? Well yes, I've already tested enough to know it is functional and produces lot's of heat, but haven't taken any measurments as yet. Will it be efficient? That remains to be seen. The concept is sound, as has been stated previously, these units are already available in Europe, and it's really just the next step beyond a dual stage water to water unit. The thing I don't know about is will my sizing of the 50 plate heat exchangers be correct, or how well those will work with the variable output compressor. No one seems to be able to address that with me.
    THIS IS AN EXPERIMENT, and perfect results are not expected, however it would be nice if something close could be achieved. I know you guys like to slap some components off the shelf together and call it a system, and would be foolish to try this sort of experiment on a customers install, but this is for me, I am willing to pay the price and take my chances on success, and have done all the calculations that are possible. My real problem is that none of the plate heat exchanger MFGs know what thier units will do in this type of application, or if they do, they don't publish anything on it and understandably, don't want to speculate.
    I was hoping that this forum had some real HVAC design knowledgable guys, and not just system designers / installers. So far the responses have shown that assumption to be false. I don't mean to critisize you guys for what you do or don't know, if I were in your buisiness I would do the same. However, you make assumptions that I don't know what I am doing with this installation that are just not true. [​IMG] If I wanted a system similar to what you might design / install, I'd just order a 2.5 ton Water furnace water to water unit and have done with it. That's not what I'm after, and not what I asked my question about. So, before anyone else jumps to the wrong conclusion that I've not done my homework, please know, I've done the heat loss calcs, I've done the ground loop and zone loop calcs, I have the proper sized circs installed, I've had the whole thing reviewed by a couple different proffesionals, and the sizing calculations for the heat pump show I need 2 tons (24k BTUs). Not only that, I have a years experience now to confirm that 4500 watts (15360 BTUs) actually provided all the heat needed for a fairly average to harsh winter. (Heat loss calc showed 11,450 btu loss at 32 degrees outside temp. Home is HIGHLY insulated ICF construction, and extremely energy efficient..).
    Now, if anyone knows anything about heat pump design and construction, NOT GEOTHERMAL / HYDRONIC SYSTEM DESIGN AND CONSTRUCTION, I'd love to hear from you.[​IMG] I'd specifically welcome comments on plate heat exchangers in freon applications and DC inverter technology in general.
  14. urthbuoy

    urthbuoy Well-Known Member Industry Professional Forum Leader

    In defense of my fellow professionals on this site, we put in systems that an average occupant can run. Ideally plug and play systems frankly. That is where I want to see the industry. There being little demand for systems requiring high education in thermodynamics or operating licences.

    If the manufactures themselves can't define a refrigerant: brine heat transfer in their devices, it is hard pressed to expect us to. It will be trial and error.

    A refrigeration forum is likely best.
  15. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    Joe may have missed the much earlier comment that you met heat load with a conventional storage electric water heater, prima facie evidence of the building load.

    You several times wrote of a "50 plate heat exchanger" While I'm able to surmise that refers to a flat plate heat exchanger there is no way we here can reliably attest to its suitability for a particular application based solely upon its number of plates - other dimensions figure into performance characteristics.

    From my (admittedly limited) knowledge of flat plate heat exchangers, 50 plates sounds way oversized for your modest load.

    With enough brain sweat and experimentation you can no doubt make any 2 ton refrigeration or heat pump appliance meet your load.

    Deep analysis of the thermodynamics involved between refrigerants and heat exchangers of whatever type (your reference to Freon is odd and quite dated, BTW.) is well beyond the scope of this site; you should have inferred that easily from browsing typical threads. Your professed disappointment at being unable to garner such information here speaks more toward your ignorance, not ours...

    Good luck with your experimentation. We look forward to hard data as to its reliability, durability and COP.
  16. Corley

    Corley New Member

    Freo.., er, eh, I mean.. Refrigeration

    You are absolutely correct on all points. I was totally ignorant regarding this forums participants limited knowledge of refrigeration technology as it applies to their professions. My bad! (Seems odd to me to be in the heat pump application / installation business and not know squat about heat pump design, but I do find that to be the case in many other industries as well, so I guess it should not surprise me???) I'll do as you suggest and take this descusion to the proper "refrigeration" forums...
    Thank you for your indulgence,

    PS RE the size of the heat exchangers, my research indicates that it is not a problem for the exchangers to be too large, and a big problem for them to be too small. Further, plate exchangers tend to clog internally over time due to scale, loosing some capcity and flow in the process. Therefore, I opt'd for the very conservative 50 plate version which is rated at 97,000 btu heat transfer which is of course much greater than my need. Physically, they are still rather small.
  17. teetech

    teetech Member Forum Leader

    Dealers do not design heat pumps.... manufacturers do.
    They have test labs, engineers, and lots of in house resources that are not available in the field. You may be interested to know that they have been testing your concept for a couple of years now. To bring this to market involves cost factors, reliability, and efficiency increases to warrant the sales.

    If you have the time and money I wish you luck!
  18. cumbelec

    cumbelec New Member

    Hi Corley
    There is a lot more to a refrigeration system than meets the eye. Here are a couple of pointers that may help and reduce the hart-ache.

    You need enough refrigerant in the system to transfer the required heat. Refrigerant just like water can only transfer a certain amount of heat, if you do not have enough you will not get the performance. You estimate the volume of both the condensation ( Heating side and original indoor air unit) and the evaporation ( Ground loop side and original external unit) and try to keep the volumes as close as possible by installing receivers or tanks before the evaporator (between the evaporator and the expansion valve) and after the condenser (between the condenser and the expansion valve). These are not the type of receiver with one pipe. The receiver must have an input and output so the refrigerant flows through. If you do this you stand the best chance of some sort of performance.

    The second thing you must do is put the sensors in the right place. If you get this wrong you stand no chance of it working. Most DC units have an electronic expansion valve so if the sensors are in the wrong place you stand no chance.

    The last thing I can suggest is you need to understand (from the operation manual) how the defrost cycle works. If you get this wrong the unit will keep going in to a defrost cycle ( effectively going in to cooling mode), generally at full capacity when you need it most. Some units measure the difference between the external temperature and the evaporator discharge temperature. When this achieves a certain differential (usually changes with ambient temperature to achieve optimal performance) it will go in to defrost cycle. The unit will be installed inside so will the external temperature will be the room temperature and the evaporator temperature will be below 30deg ( below the loop discharge temperature) therefore the software will think the external unit has frosted up and change to cooling to defrost. I suggest you install the external temperature sensor on the loop flow pipe. This will keep the temperatures closer and therefore confuse the software.

    You will not need a buffer tank if it works.

    I wish you the best of luck. :D

Share This Page