many of the geo systems we have been asked to interface with in the past had cycling related issues. that is, once a cycle was over, the unit has to time out before it turns on again. the ECR system doesn't seem to care, from what I"m told, which should allow us to fire it up, using an 80 gallon buffer, to maintain a RESET water temperature in the tank without concern by using a smaller firing temperature differential. during lower load conditions, it may cycle more often. is this a big deal or not, and is it not a big deal with the ECR unit for a particular reason? can other units deal with occasional "shorter cycles" (not quite short cycling) without problems? not having to run a 20 degree differential to ensure long enough run times and storage means we can skip mixing...

Starts are hard on compressor motors, and short cycling doesn't give refrigerant time enough to equalize within the system during the off cycle. When a compressor tries to start against unequal pressures it generally trips out on an internal overload. That said, a 20 degree delta across 80 gallons is 13,300 Btus, well over a ton-hr - seems like more than enough buffering to overcome the problem

yes, but the goal is to greatly narrow that delta t (to 5 or 10) to better provide the necessary water temperature to the system directly, instead of a swinging, wild approximation across such a wide dt.

and preferably, without quadrupling the size of the buffer.

The ECR guy I was talking to didn't seem to think that cycling issues existed with that unit (earthlinked). Is he wrong (very possible) or is there something special about that unit that overcomes this cycling issue?

Rob, great topic. I think there's a lot of room for improvement in the way hydronic geo systems are typically plumbed and controlled. In general I think fewer compressor cycles are better from a reliability and efficiency standpoint. Extreme short cycling can kill a compressor regardless of the system it's attached to. Maybe the DX system allows the pressures to equalize faster and therefore requires a shorter off-cycle, but even so I still wouldn't want to see a system cycling every few minutes without some long term reliability data to back it up. How short a cycle are you talking about? The internal lockout on my w-w unit requires about 4 minutes between cycles which doesn't seem too bad.

I think this should be a moot point, though. I think it's possible with the right design to avoid short-cycling without a huge tank. My own system(which I am still improving) uses a 120-gallon tank for a heat pump that typically puts out 60-80kbtu/hr, so the tank is 50-100% oversized based on industry standards. I control it with a Tekmar 260 (slightly modified) and this last winter it kept the house temperature within +/-2deg even though I never got around to hooking up indoor feedback. That's with constant circulation and outdoor temp feedback ONLY - no thermostat or indoor sensor. It typically cycles about once per hour, the run time varies with outdoor temp but the number of cycles per hour stays relatively constant. I do get a temperature "spike" going to the radiant panels at the start of each cycle which I'm looking at ways to reduce without doing a lot of mixing. It isn't noticeable in terms of comfort, and I wouldn't even know it was happening if I wasn't logging temeperatures. My only concern would be that it might cause some noise issues in lightweight plates. As I said it's a work in progress, and I think once I get it optimized the buffer tank will turn out to be bigger than necessary.

I can send you more info and data later on if you're interested - got to do some "real" work now!

I would LOVE to see that data, and any commentary you've got on it as well.

the trick is, we are seeing from our ECR guy that COPs really change when you are producing cooler water. so to aggressively lower water temps (to 90 or 100 degree maximums sometimes) we are running into dt as an issue. for instance, if I want to produce a max of 100 degrees, but I have a 20 cycling dt, my average water temp is only 90 degrees.

I could raise the system water temp to 110, leave the dt alone... and lose almost a full point of COP, I think... at least that's what the data this fellow seems to indicate (though I'm not positive his methodology was flawless).

if I could lower the dt to 5, my max supply temp goes down to 102.5 if I need an average 100 degrees.

see what I mean? but there may be other holes in my logic as well... ultimately, it's all about COP.

I understand completely what you're saying about the average water temp vs what the heat pump is putting out. This is not unique to DX systems - w-w system COP has the same strong dependence on outgoing water temp. It is amazing to me that systems are commonly installed without any reset control - they just set an aquastat to 115F or whatever they need on a design day and let it run all year at that. You can easily lose 25-20% efficiency this way.

My goal is to keep the buffer deltaT the same as the heat pump deltaT - you can't do any better than that. The temperature rise though the heat pump is more or less constant, though it will vary somewhat as the loop temp and hydronic temps change. So, your best-case theoretical heat pump leaving temp will be average radiant temp required + heat pump dt/2. If you buffer control dt and your radiant dt matches the heat pump dt you have the best case for efficiency. In practice this is hard to achieve, but I think you can come pretty close if you can maintain stratification in the buffer tank.

I'll try to send you something later today - I need to clean up my plots so that they are understandable.

that's an interesting thought. how would I nail the heat pump dt (water side)? I imagine it varies based on exchanger, water temp, flow rate and collection field size.. is this something I can pull from MFG lit somehow, or is this one of those "set it up and fiddle" things? or.. does it end up being fairly constant in the field amongst different brands of units?

so much to learn :D thanks for the help though!

Rob, the temp rise or heating capacity should be available in the manufacturer's literature. Generally the flow rate is pretty constant, especially if you're using a buffer tank, so the temp rise is just capacity divided by flow. You should be able to get a reasonable estimate of flow from the manufacturer's pressure drop tables (I know in your case you'll be actually sizing the pumps properly rather than slapping the recommended # of 26-116s on!).

Assuming the flow is constant, the temp rise is proportional to the heating capacity which varies both with loop temp and hydronic temp. Here are 2 plots from my system - temperature rise vs time over the last few months and temperature rise vs loop and hydronic temps - you can get an idea of how much it varies. From what I've seen 10-15deg seems to be the industry norm.

I haven't forgotten the data on buffer tank temps, I'll try to put that together soon. Busy weekend!

Just to make sure my head is wrapping around this properly; amongst all this variance, electrical consumption of the unit is constant, yes?

So these varied dts are actually representing a capacity/COP change, based around a set kwh consumption rate?

Just making sure I understand.

Actually, the power consumption of the compressor varies a great deal. For a hydronic unit there is very little dependency on loop temperature, but a strong dependence on hydronic temperature - going from 85F to 125F leaving temp can almost double the load. So as the hydronic temp rises, the capacity declines and the power consumption increases, so the effect on COP is multiplied.

It always has to be complicated heh... what are the different plots on that compressor power graph? I get that the colored dots are ground loop temps, what are the lines under that?

and thanks for the all the info!

Sorry, should have labeled that better, I just grabbed an old plot. The soild lines are the power spec from the manufacturer at different loop temps - as you can see my unit draws more power than it should, but that is a topic for another time. Mainly I just wanted to show the variation in power with temperature. Most manufacturers have this data available in their specs (and if they don't they should, imo).

Guys - trying to follow your thread - what is reset control? Do I need one if I have geo but am not going hydronic?

Reset changes the temperature of something based on something else. it's "temperature reset".

generally you are varying the temperature of a radiant circuit with a mixing device, or a tank, or a boiler, and generally it's based on outdoor temperature, so it "asks" for higher water temperatures when it gets colder out and your heat load increases.

not sure how it would work in non hydronic situations, sorry.