Has anyone ever used room t-stats instead of buffer tank temp to control a hydronic geo system? I've never heard of doing it this way, so tell me why this control scheme wouldn't work. I'm imagining a low mass radiant system for a very tight, well insulated home. I understand most water to water systems use thermostatic control with a buffer tank, sometimes including an outdoor reset to lower tank temperature during mild weather. Has any ever let the tank temperature float? What I'm thinking is a two stage (house/room) thermostat. In first stage, the circulator pump turns on and circulates water at whatever the tank temperature happens to be. If call can't be satisfied, t-stat calls for stage two. At a stage two call, the geo system is engaged. Once stage two is satisfied, t-stat drops to stage one and geo shuts down. Water continues to circulate from tank and may or may not finish the call. When the geo is engaged, the flow would be: 1. From the tank. 2. Through the Geo HX 3. To the floor. 4. Return to tank. Tank temperature is thusly increased gradually by the return flow. (How gradually would depend on tank size and Delta T on the floor loop, you know, engineering..) Benefits I see are these: Geo is always being fed lowest possible temperature on load side. Tank temperature is determined by load (via run time) rather than a set point. Plus, a fairly simple (inexpensive) control system.
Compressor cycles I think the biggest problem with that control would be compressor cycles too much. It's basically the same reason the Bristol twin single's burnt out. The thermostat starts on first stage, but can't keep up. Second stage is energized and the space warms close to set point. Stat decides it's close enough and drops back into first stage, home can't keep up and stat engages second stage again. Turned out to be very bad for the compressor. Bergy
Sorta get it I may be a bit confused. But geo out to the floor will create varying flows/temps in return to the unit. The units are not a big fan of this. Could be corrected with variable speed I guess. What is the thermostat going to be measuring to dictate stage 1 or 2? The return temperature on the load side? This may be a poor representation of water temps if zones are opening and closing and different water temps return from different areas. The buffer would give a more even representation.
Varying flow through the unit: yes - if zoned, and O.K., most people zone radiant. In the context of a super efficient shell, though, you might be able to get by with one zone - which is the application I have in mind. But if zoned, as long as every zone meets the minimum flow requirement for the unit, why would would the unit care if it saw varying flows? Even if zoned, though, the entering load temperature would vary only slowly as the buffer tank would "average" out the returns plus the mass of water in the tank. So no matter how many zones were operating, entering load temperature of the unit would never see any "shocks". Over the course of an hour or two, the temperature would change, but not over minutes. The t-stat is not measuring water temperature at all, simply room temperature. The differential between stage 1 and stage 2 would be small enough that the room occupants would never sense the difference. I can't see how it would cycle any more than forced air system, and probably less because there would still be stored heat circulating with a stage 1 call, just at a lower temperature. This might actually benefit a super shell house in that there would be less chance for overshooting. The main reasoning here is that with geo, load temperature matters a great deal for efficiency. It's not just the BTU's that the house needs, it's how warm you have to make the water to deliver those BTU's. This is an attempt to see if one could deliver those BTU's at as low a temperature as possible without overly complex controls. In a nutshell, the longer the system runs continuously (in multiples of hours) the warmer the buffer tank eventually gets until it reaches the maximum design temperature. Otherwise, if you can get enough BTU's into your house on a mild day with 85 degree water, within a few degrees, that's what your tank temperature would be.
Then I would just keep plumbing as normal, and vary the period in the temperature drop. Allowing it to cool a lot more before heating.
If I'm understanding, something like letting the tank swing between 85 and 110? Geo on at 85 and off at 110? Other than always "finishing" the tank to 110, I think it would be functionally similar to what I've outlined - especially when looking at how the system would behave when starting with an 86 degree tank. The difference (and possible room for efficiency gain) is the "next call" after the tank reaches its 110 degree high mark. At that point the circulators are pumping 110 degree when the load might be satisfied at a lower temperature. However, with a relatively small buffer tank, the hit in efficiency is probably minimal.
A slightly different take O.K. - with the input here, here's how this could be simplified further. Use common plumbing practices i.e. zone circulators fed from buffer tank with geo "charging" buffer tank on its own circuit. But rather than queue up the geo based on buffer tank temperature, use room stats to call the shots. (I don't think this would work very well with a slab, the lag time is too much) Room stat (or zone controller) stage 1 call = circulator pump(s) only drawing from buffer tank. Room stat (or zone controller) stage 2 call = geo operates charging buffer tank until call met. Room stat (or zone controller) stage 3 call = electric resistance boost (with this design, boost would be in HRV duct) Stats would be set with 0.5 degree differential (and/or zone controller set for fairly aggressive upstaging) between stage 1 and 2 for comfort. So now the question might be, why not just go forced air? Well, other than wanting radiant, what I'm really trying to do is heat two low-energy (and low income) duplexes with a single geo unit to save installation costs. Since I can't interconnect ducts between units, I'd have to go water to water. I'd use the same unit for all domestic hot water in both units as well. And because you guys are sharp - I'll add that I'm trying to get the loads down to a single 2 ton unit, so loop size really isn't much of a cost factor.
Blake, among the problems with your premise is we are contractors who swore a comfort oath. You can do a duplex with a two ton as long as your load is under 40K heating in my neck of the woods and cooling is under 30K. Tennants don't tend to be fans of droop or 78* in summer.
I can easily cover the heating loads with 2 tons. Cooling is another problem - it's not tonnage, it's delivery. With the current plan, the only ductwork in the house is for the HRVs - not many CFM. The plan is to run two cold water coils on the fresh air inlet - one upstream and one downstream of the HRV unit. Between two HRV's I figure I can get about 1 ton cooling (latent +sensible). I understand the sentiment of comfort at any (sometimes a lot ) cost. However, for workforce housing units that rent for 1,100 a month including utilities, I'm not looking to to provide luxury. I'll ask my root question a different way: Has anyone ever used room stats instead of tank temperature to control a hydronic geo system? (I edited my initial post to pose the same question)
Hello Blake Guess what I do what you suggest all the time, every time. I have found a series of controls that will allow the heat pump to satisfy a heating or cooling call to the buffering tanks directly. Joe Hardin and I looked at such an issue last spring. I use a lot of water and so water to water geo heat pumps are my favorite way to make things work. What I do is use a buffering tank as a giant hydrolic separator. See the Caleffi Idronics series on their web site. They are free down loads. A master thermostat, (think heat pump capable stat), determines heating or cooling. This could also be done with an out door re-set type of control. The geo heat pump make a full tank of water, (warm or chilled),then the zones draw their needs based on cheap stats that can do heating and cooling. The cheap stats just bring on pumps or open zone valves with variable speed pumps. The user interface is what they know, a wall mounted thermostat. It responds to air temperatures like they are used to having happen. I always use room stats to control my systems. The buffering tank(s) then power the calls while the heat pump(s), and this system allows one to leverage the net BTUH available from a machine. We are taking a few days off from such a system and are camping in the wilds of Ohio. I love the ability to say call me while I try for spring Bass. I have spent years trying to teach heat transfer equipment to do what I want it to do, not what is the norm. If I can help give me a call@440.223.0840. Good luck, Mark
low cost control system Many great points have been brought up about the flaws in trying to control a heat pump with a room stat, so all I will say is just put in an HBX CPU-0500 http://www.hbxcontrols.com/Resources/marketing/cpu-0500.pdf they can be had for $400 something, doesn't get much cheaper or simpler than that. When set accurately it will track the heat requirements of your house perfectly and your heat pump will be very happy.
Here are a couple setup examples: Temperature and Energy logging by: Web Energy Logger Variable speed with zones valves. No outdoor reset since the the building is so efficient that the buffer tank is 85F max. A zone opens up, buffer tank temp drops down, heatpump turns on. Simple. Less efficient houses requiring higher load temps, so they have outdoor reset Web Energy Logger: Here thermastatic valves at the radiator create the zones. Then you can stack multiple heatpumps for higher load, also with outdoor reset. Temperature and Energy logging by: Web Energy Logger The principle is always the same.
