I posted this in the geo forum by mistake.

Here is the issue with limited information.


The house is new ICF construction, 5T water to water Geo unit, have a 50 gallon buffer tank, lower level floor is concrete, main floor is gypcrete. System was designed for the house or so I was told. There is NO outdoor reset. Floor tubing is in the gypcrete and concrete 12" on center.

On cloudy days, the house stays comfortable, just on sunny days when the geo does not run all day is when there are comfort issues. The house temp stays above the set points on the thermostats and the water in the system cools to about 82 degrees. The normal set points on the system points are geo on at 92 and off at 98. So when the whole system is calling for heat and the water temp is 82 the geo runs all night before normal temps are regained.

So, slow to heat thermal mass with slow to heat water gives me uncomfortable house on cold evenings after a sunny day.


Any thoughts from the pros?

You need to use the high mass/slow reacting floor less and something faster reacting more. How about a water-air heat exchanger + fan?

Are you saying this is a bad application for GEO radiant?

Forgot to mention location is Wisconsin.

Floor sensing can help make the floor not get so cold during off periods, which reduces the lag time coming out of shutdown significantly.

also, intelligent controls, we use Teknet 4 from Tekmar, can anticipate better than most and change water temperatures as well to react properly.

I agree he has control problems.

Do you think tube spacing is adding to this problem?

I do, it will slow reaction time unless he's truly maxed out his geo capability in the "recovery heat dump period". However, good controls could minimize that as a problem by reducing how much "catch up" he has to play in the first place.

You could also increase the size of the buffer tank and heat it (but not the floor) up during the day (when you have excess capacity) and discharge it in the evening (when you are limited), preferably directly to the air. But very much mass in a heating system (vs in the home) that can't be disabled isn't a good thing and increases unwanted fluctuations.  

I do, it will slow reaction time unless he's truly maxed out his geo capability in the "recovery heat dump period".

Well we don't know the loads so guess work is all we can do.
By specs he probably has about 50,000 BTU for recovery.
It's a HP not a mod-con so that's what you design for.
Manufacturer recommends at least 80 gal buffer tank.

It all goes back to design and controls play a big part but I'm having trouble with 12 Inch pipe spacing with GEO.
I also wonder if water temp drop is excessive for 8 hrs.

If he can maintain room temps at 95 degree water with the geo system normally, then the 12" spacing is not a problem. If he can never maintain room temps and the heat source cycles off and on, then you need to raise your water temp. If he can never maintain room temps and the unit never shuts off, then it's overloaded. however if you're playing "catch up" it will act overloaded until you do, in fact, catch up, meaning the mass is up to temp again.

I hope that helps but naturally it's pretty tough to do a real trouble shoot in a forum.

I'm not a pro but why wouldn't the system be set up to keep the buffer tank hot all the time? Otherwise you have to heat up 50 (or 80) gallons before you can even start putting heat to the floor. So instead of acting as a buffer tank its acting as an additional uneeded load.

Also, why not an outdoor thermometer to sense the cooling evening temps and kick in the circulation in anticipation of the nightime load?

> Also, why not an outdoor thermometer to sense the cooling evening temps and kick in the circulation in anticipation of the nightime load?

It might help some if you put the sensor in the sun (so it could detect the real issue of solar gain). But you can't get around the generic passive solar heating problem that you want little or no heat output at 4PM and then full output in the evening. A high mass system cannot change output that fast. Similar problem in the mornings - you need full output (say 85F slab) to be comfortable and then suddenly the clouds leave or the sun comes over the hill and within minutes you want to shift to no output (70F slab).

The control for the unit is a water temp sensor at the buffer tank maintaining 92-98F.
The wall stats control water pump and 4 zone valves (flow).

Given that, it sounds like a slab sensor/s (for flow) is his best option to keep floor from getting too cold.
There will probably have to be ODR also to accommodate weather changes (or load)

Do you agree?

you can replace heat as you go or you can wait until you need it.

If you wait, standy losses are lower, cycle times are longer, and your initial geo efficiency is higher because the water temp is lower.

However in the systems we use (typically teknet 4 to aggressively lower that water temp) in practice they are nearly constant circulation and so the tank is basically maintained anyway. but IF all the zones shut down for some reason, it won't maintain the tank.

some kind of reset and floor sensing will be a very good start. indoor feedback plus ODR plus floor sensing will be as close to perfect as you can get in this configuration.

i have almost the identical setup, and don't see as big of heat swings he's seeing, but I agree, outdoor reset is a necessity and we'll be adding that to ours this summer.