Buffer Tank Size vs COP when Cooling

Discussion in 'General Discussions' started by Altnrgy, Jun 18, 2012.

  1. Altnrgy

    Altnrgy Member

    Scenario 1: A large (~300g+) buffer tank that rarely gets above 105f even on the hottest days. It spends most of it's time under 95f. Two person household thus hot water usage is relativity minimal. Even at 95f that's still 30 less degrees the water heater has to add to the 65f well water.

    Scenario 2: A modest buffer tank that daily hits the 125f DSH set point and the DSH pump shuts off. Assume same two person household.

    Which of these theoretical scenarios purchases the lowest total cooling and domestic hot water kWH's during the summer? This is just a total kWh question so ignore the economics and the space required for a 300g buffer tank.

    I think Scenario 1. Based on my assumption that the ground loop is seeing lower return temps because it's always able to reject heat to the buffer tank. The less heat that you reject to the loop the higher COP of the GEO unit. In this particular household the HVAC load is much higher than the domestic hot water load so it makes sense to optimize towards the HVAC load.

    Am I approximately on the right track? I realize I could be splitting hairs.

    If you kept both scenarios the but added 3 teenagers to the household now the scenario might change. Or would it?

  2. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    With modest daily hot water use, the 300g tank would eventually warm above 95-105 simply because its capacity likely represents several days' hot water use, so the water in it would have several days to be heated before being used.

    A 300g tank could have substantially higher standby losses owing to large surface area.

    A 300g tank will likely cost a king's ransom since (I believe) any pressure vessel above 120 gallons must be ASME rated, adding greatly to weight and expense. That could be a $5k tank.

    My general advice (but know that some here with as much or more knowledge than I disagree) is to size the buffer for a DAY's hot water use and the finishing tank for the highest HOUR's hot water use (3 hours if heat pump water heater)

    Some common sense needs to be applied matching buffer tank to system tonnage...regardless of use I doubt I'd pair a 2 ton system with a 120 gallon tank nor a 5 ton system with a 40 gallon tank.

    I have a 3 T system locked in low stage, an 80 gallon preheat tank, an 80 gallon HPWH final tank, 2 adults, 1 teen and 2 preteens. Buffer tank runs around 100-110 most of the summer, final tank uses about $5 in power in summer, $10 to $15 in shoulder months, $5-10 in "winter" (Florida)
  3. Altnrgy

    Altnrgy Member

    All of that makes good sense but I guess I didn't make it clear that I'm not talking about weather or not it's feasible. I already know it's not feasible.

    What I'm wondering is that if total system efficiency is better when the buffer tank is so large that the DSH always has a place to put heat.

    The theory is that the more heat you put in a buffer tank the less you put back in the ground loop which of course means your unit is running at a higher COP. And I'm only talking about cooling season.

    So does this higher COP save more kWh's than what you have to put back in the water heater to get it the final temp.

    As you point out there are a lot variables that could affect this which is why I have the two scenarios which were at opposite ends of the spectrum.
  4. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    DSH is not guaranteed to "always have a place to put heat" merely by enlarging the buffer tank.

    Assuming buffer is insulated well enough so that standby losses are minor compared with hot water demand, the only way to guarantee DSH "always have a place to put heat" is to consume enough hot water so that buffer tank temperature is always below compressor's refrigerant discharge temperature.

    DSH diverts, on average, no more than 10% of system's total capacity. Tweaks to the DSH and its buffer tank will have negligible effect on ground loop operating conditions.

    The desuperheater does just that...desuperheats. Refrigerant superheat at the compressor discharge is high quality but low quantity. Stated another way, the refrigerant temperature is usefully high for purposes of heating domestic water but the actual quantity of heat available at usefully high temperature is small. This arises from the low specific heat of superheated refigerant relative to the much higher heat transfer associated with refrigerant phase change, that is, the transition from gas to liquid and back again.

    Much, if not most, of the excess superheat at the compressor discharge originates from the compressor's internals; mechanical moving parts and motor windings.

    My advice on buffer and final water heater tank sizing is based on the goal of bridging the daily gap between spikey domestic hot water consumption and long slow desuperheater hot water generation
  5. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    Remember in the heating season, if you have satisfied the buffer tank, your space heating capacity improves.
  6. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    "satisfied the buffer tank" is an interesting turn of phrase since the buffer lacks any thermostat or similar control intelligence.

    As buffer tank temperature approaches compressor hot gas discharge temperature, heat transfer / diversion falls well below the nominal 10% typically cited for a 90*f buffer.

    If by "satisfied" you mean that the buffer gets so hot so as to trip the antiscald high temp lockout, then diversion drops to zero. I'm not intimately familar with all the refrigerant temperature / pressure / superheat / subcool conditions at all vertices of a PT chart, but I suspect that a system working off fairly chilly (30s) loop water offers scant compressor discharge energy hot enough to trip a buffer DSH pump antiscald interlock.

    If nothing else, I figure there is a Peter / Paul effect in play...if a bit of unwanted diversion of heat occurs on a cold day from space heat to buffer tank, the transferred Btus will offset finishing tank water heater energy use by a like amount.

    I know there is a possible issue with diverting heat from the conditioned volume to a tank in an unconditioned basement, but I suspect the effect is minor since a home atop a basement at least partially benefits from any extra heat provided there.
  7. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    It depends a bit on the size of the unit and the run time, but I alway prefer a smaller tank. Quicker response time, hotter water temp, the second tank does not have to make up so much heat.
    With 2 people, there is really not enough water usage to justify a larger tank. Not sure if I would change that strategy if a couple kids were added. 50 gal buffer tanks run well in Western New York during the heating season, cooling DSH output is bad due to 65F max loop temp.
  8. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    That's a fair point.

    My concern is that if most of a household's hot water use is concentrated around one end of the day (in our case 80-90% of hot water use occurs between 7 and 10 PM), and the buffer is sized much below that volume of concentrated use, the finishing tank gets a slug of completely untempered water at the tail end of the heavy use period.

    Suppose a family uses 70 gallons during the PM rush. A 50 gallon buffer might supply 50 gallons at 120*F folowed by 20 gallons at 60*F. Is that better or worse than an 80 gallon buffer supplying all 70 gallons at 105*F?

    That math may work out to a wash.

    I suspect that there may be a bit of South / North regional differences in play around this question. Down here, cooling-dominated, we have to err on the side of caution with system sizing. I do my best to talk clients down a half ton or so, but fundamentally, there is no such thing as a "10 kW / 34k Btuh aux COOLING strip" to bail us out during the 10-20 or so really hot days when temperatures exceed design.

    If I could size compressor sections to meet just 90-95% of the cooling load, I'd sell a whole lot more 2-3 ton systems than 3-5 ton systems.

    So what does that have to do with hot water / DSH buffer tank sizing? Good question...glad you asked....my answer considers the fraction of high vs low stage (hereinafter Y1 / Y2) operation. In Y2, the compressor discharge temperature is significantly (20+*F) hotter than while in Y1. That means that if a system spends significant time in Y2, say 2-4 hours per day, it has the potential to make a moderate amount of very hot water (130*F+). It may make sense to concentrate that high quality hot water in a smaller buffer tank, completely precluding finishing tank firing.

    OTOH, if a system sees relatively low Y2 operating hours but loads of Y1 time, it'll make large quantities of lower quality hot water, say around 100-105. Refrigerant discharge temperature in Y1 is such that after the buffer reaches 110*F or so, little or no more heat transfer takes place. Therefore it makes sense to upsize the buffer to capture more of the low grade heat, knowing that the finishing tank will always have to add 10-15*F, but keeping it from getting cold slugs needing 50*F rise.

    I hypothesize that a properly sized system up north experiences a Y1 / Y2 fraction of perhaps 65 / 35, since it is undersized a bit for design heating load. OTOH a southern system selected to meet 99% design temperatures without assistance experiences a Y1 / Y2 fraction of 85 / 15 or even 90 / 10.

    Futhermore, there are north / south differences in hot water utilization. Northern households use less hot water per person per day, but they heat it hotter. A study I have indicates northerners use 16-18 gallons of hot water / day but the typical setpoint is 140*F. The study hypothesizes that northerners shower less often (once daily), but water need be hotter owing to greater piping losses.

    In the south, a walk to the mailbox makes us break out in sweat May through October, so we shower more often (twice daily), but water can be cooler owing to higher ambient wet and dry bulb temps and lower piping losses.

    Executive summary:

    1) Y1 makes DSH warm water, Y2 makes DSH hot water
    2) Northern systems operate much more often / longer in Y2, able to make hot water
    3) southern systems operate mostly in Y1, so can make mostly only warm water
    4) Northerners need hot water, southerners need warm water (but more of it)

    5) Southerners are better served with large warm buffer tanks; Northerners better with small hot buffer tanks.
  9. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Add here that the heat pump runs the most during the night, prior to the most host water usage, whereas the run time is the least during the cooler night in A/C dominated climate.

    It is certaily more complex than "bigger is better" or "smaller is better". Now add 3 teenage daugthers or " me and my dog" usage into the equation, and different loop temp, or the question if the geo system would have a W-W or W-A heatpump, and the resonings wiill change again.
  10. AMI Contracting

    AMI Contracting A nice Van Morrison song Industry Professional Forum Leader

    DJ beat me to the night-time addendum.
    Otherwise a fair dissection of different buffer designs vs climate.
  11. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    I agree that a cooling dominated climate experiences paeak in early PM whereas a heating dominated climate experiences peak in early AM.

    Consider the hot water peak. Some families may all shower during morning, so a heavy DSH contribution during the cold predawn is especially helpful to mee the AM peak.

    If peak hot water use occurs in evening, then there is a delay up north between production and use. That delay is why I favor larger buffer tanks - ability to store more heat and carry it forward.
  12. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Larger tanks simply don't get hot enough in A/C mode up north due to limited output in A/C mode with R-410a units and looptemp maxing out at 65F. Right now my loop is at 55F, not much desuperheat. So the second tank's electric heat element will run more if the the buffer tank is too large. You find that out pretty qucikly if you put the system on a WEL and monitor it.

    Temperature and Energy logging by: Web Energy Logger
    Here is another example, it is getting warm here, but 50 gal bufferr tank rarely gets about 75F. Bigger tank would perform worse.....
    Last edited: Jun 29, 2012
  13. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    I get it.

    However, I still maintain that DSH works best if buffer volume approximates a day's hot water use.

    Suppose up north a 50 gallon buffer tank warms to 75, and an 80 gallon buffer warms to 70. With the 80, the finishing tank rarely sees raw cold water at 50-55.

    Another viewpoint to consider is that DSH output (in other words, heat transfer) is maximised when delta T between refrigerant and water is greatest. Since a bigger buffer runs cooler, it gathers more heat.

    Whether the minor performance increase is worth the extra cost and space of the bigger tank is debateable.
  14. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    I had the same thought, especially with multi heat pumps systems feeding in one buffer tank. Tried it. Did not work well. Even with smaller buffer tank, the second tank rarely sees 50 degree water, since the response time is much quicker. Again, only valid for our climate up north.
  15. Altnrgy

    Altnrgy Member

    Please allow me to restate my original question but instead lets compare two identical systems closed loop systems with one exception. One has a buffer tank and the other does not. And also let me attempt to make it clear that I'm only interested in the COP of the unit and am not talking in shape or fashion about optimizing or even proper buffer tank sizing. This is just a COP question.

    During cooling season would the COP of the unit with the buffer tank be higher? It has to be because there is less heat being put back in the loop which equates to a lower EWT. Right?

    So continuing with that thought would it not be true that the larger the buffer tank the higher the COP of the unit would be? This is still just theoretical, just during cooling season and has nothing to do with whether or not the buffer tank is properly sized.

  16. engineer

    engineer Well-Known Member Industry Professional Forum Leader

    OK, but clarify whether the system without a buffer tank is also without a DSH heat exchnger?
  17. docjenser

    docjenser Well-Known Member Industry Professional Forum Leader

    Yes, the COP would be higher with a buffer tank because heat gets diverted into the tank instead of the loop, lowering your EWT. But at the start of the heating season, your COP would be lower since in heating mode you COP is now lower. Plus, how do you make your hot water? It is literally free in the summer, except for the small circulation pump pushing it into the tank.<br />
    Curts argument is that a larger and colder tank increases the delta T between the refrigerant and the load temp inside the DSH heat exchanger, this increaseing the amount of heat which can be captured. Which is true, but I counter with the argument that now the water feeding into the final tank is colder, and electric resistence has to make up more heat.
    Two different philosophies which reflect two different experiences with the completely different operational patterns. Curt is down in Florida, I am up north in Buffalo, New York.<br />
    He starts of with 70-75F degree water coming in the DHW tank, we have 50F all year around. He runs things in A/C mode most of the time, so he has a lot of heat to get rid of, I do not. For me, 90% of the run time is in heating mode, where the DSH make twice as much heat, but also runs much hotter, but now steals away COP and capacity from the heatpump.
    I could go on and on, and it gets more complex. At the end of the day, the DSH is a cheap way to make a large amount of your hot water. But it is not perfect, since it is not on demand, and it requires a buffer tank to run efficient. Without a buffer tank, don't even bother.
    Last edited: Jul 1, 2012
  18. Altnrgy

    Altnrgy Member

    Yes, with a DSH vs. without a DSH.

Share This Page