Posted By jonr on 04 May 2013 07:39 AM
Outdoor reset controllers allow you to put in whatever margin (extra temperature) you need for rapidly changing conditions. So you have extra capacity immediately available and can use it for heating up the houses's thermal mass. A smaller tank doesn't provide as many btus to do this (ie, is MORE likely to fall behind that a properly configured system with a larger tank).

For example, rapidly changing outdoor conditions mean that a concrete floor slab needs to rapidly change from 80F to 85F. What is better able to do that - 50 gallons of 95F water or 500 gallons of 95F water?

How do you get the buffer tank to 95 F to begin with? You have to run the system at 95F before it gets colder! Running the system at 95F instead of 85F will render the system about 15% less efficient. Besides running it at higher supply temps, it reduces the capacity of the heatpump, now you more likely need a larger unit. Again, a buffer tank is not a storage tank, its purpose is flow separation and to prevent short cycling.

You have to run the system at 95F before it gets colder!

Yep, but for equally fast response to changing conditions, a small tank would have to be far hotter than that! Even less efficient! If you decide that you prefer efficiency over comfort/fast response, then operate the tank right at what current conditions require. If you plumb it right, the HP will meet the temperature requirements of the radiators before it starts increasing the buffer tank temperature. That's right - you can send 95F water to the radiators while the buffer tank remains at only 80F. Or you can have separate tanks for buffering and storage, but that's more complex and unnecessary. But those are subjects for a new topic.

Besides running it at higher supply temps, it reduces the capacity of the heatpump, now you more likely need a larger unit.

This simply doesn't apply. Design day always requires higher supply temps and if your HP can't produce design day btus at the required supply temps, then you need supplemental heat (probably not a bigger geo HP).

Posted By jonr on 04 May 2013 10:44 AM

You have to run the system at 95F before it gets colder!

Yep, but for equally fast response to changing conditions, a small tank would have to be far hotter than that! Even less efficient! If you decide that you prefer efficiency over comfort/fast response, then operate the tank right at what current conditions require. If you plumb it right, the HP will meet the temperature requirements of the radiators before it starts increasing the buffer tank temperature. That's right - you can send 95F water to the radiators while the buffer tank remains at only 80F. Or you can have separate tanks for buffering and storage, but that's more complex and unnecessary. But those are subjects for a new topic.

You get the response time by sizing the HP to the design temps or larger, that way it can inject more BTUs into the radiant mass. That way you have the comfort and the efficiency. Only at design temps you are getting close tho a steady state. One reason why you make usually make sure that a hydronic HP can carry the entire load, and slightly more.
I like to see what you define as "plumbing right". Normally I would chuckle about this, but I am actually building a system where you purposely go from the HP through the zones and buffer tank at the same time, with the same temp going into the zones and buffer tank. So how do you run the zones at 95F and keep the buffer at 80F? I would be very curious to see how you accomplish that and ensuring enough flow through the heatpump when only one zone is calling? Specifically with a 15F degree delta T!

Besides running it at higher supply temps, it reduces the capacity of the heatpump, now you more likely need a larger unit.

This simply doesn't apply. Design day always requires higher supply temps and if your HP can't produce design day btus at the required supply temps, then you need supplemental heat (probably not a bigger geo HP).

Why does that not apply? First, design day does not always require higher supply temps. We have radiant systems designed without outdoor reset, where 85F is the buffer tank temp under any condition. It is significantly more economical to increase the HP size to cover the entire load than to install and maintain a dual fuel solution.
The only time I did use an electric element in a buffer tank as the supplement heat stage when I was missing a few thousand BTUs to cover the 100% load day, and the larger heatpumps (6+tons) did not qualify for tax credits. Thus it was more economical to have a 6 ton plus electric heat element in the buffer tank installed, which comes on at close to design temp, than go with 2 smaller heatpumps to carry a 7 ton load.

First, design day does not always require higher supply temps. We have radiant systems designed without outdoor reset,

OK, if it wasn't clear to you I'll add words: "design day always requires higher supply temps than what non-design days require". If you want to ignore the temps required for low loads and run something higher, that's up to you and the customer that will have to pay for the decreased efficiency.

What's your average price for an extra ton of geo (HP and loop)?

I am actually building a system where you purposely go from the HP through the zones and buffer tank at the same time, with the same temp going into the zones and buffer tank.

You may be on the right track. Now think about small differences in pressures and which way flows will go under various conditions.

I must agree with Doc,no matter what type of cooling or heating plant there is a huge difference between a buffer tank and thermal storage.Thermal storage only works when one can take advantage of off peak utility rates.Buffer tanks are only needed when the volume of the piping is not sufficient to prevent short cycling of equipment.Typically 6 gallons per ton is what most chiller mfgs recommend.Very rarely do you find a buffer tank on a commercial system do to the well designed piping engineering.

Posted By jonr on 04 May 2013 06:57 PM

First, design day does not always require higher supply temps. We have radiant systems designed without outdoor reset,

OK, if it wasn't clear to you I'll add words: "design day always requires higher supply temps than what non-design days require". If you want to ignore the temps required for low loads and run something higher, that's up to you and the customer that will have to pay for the decreased efficiency.

What's your average price for an extra ton of geo (HP and loop)?

Retail HP price according to WF residential price list
4 ton (NSW 050) = $7712
5 ton (NSW 060 = $7902
6 ton (NSW 075) = $7949
Your loopfield do not have to be any larger for a ton more to cover peak loads during a short cold span, since you are not extracting significantly more heat out, you simple increase response performance.
I am not ignoring the temps needed for low loads, but everyone who has designed and built those systems successfully knows that you cannot go much below 85F supply temp to get any useful response time and heat transfer out of it. So you think a customer who has a radiant system designed for 85F supply temp will be penalized by decreased efficiency? I hate to tell you, those systems perform very well. Weren't you the one who wanted to heat up a 500 gallon tank to 95F?
Jon, forgive me for asking, but have you ever put radiant systems in or designed them, and had to ensure that they work well for the customer? You sometimes have good thoughts here, but you seem to completely miss the experience to know what works well in the field, and what does not? And every time I ask you how to address a problem raising out of your suggestions, you don't give any answer. Like how you run the radiators and the buffer tank at different temperatures. What is you experience and background with geothermal and radiant?

Your questions are getting ridiculously rhetorical and as I suggested, if you want to discuss buffer tank plumbing, start another topic. My experience is clear - click on my name. And read where I suggested getting a review by a good hvac engineer (more installers need to do that).

I agree, practical experience is valuable, as is an engineering degree and a good understanding of thermodynamics and hydraulics.

you cannot go much below 85F supply temp to get any useful response time and heat transfer out of it.

You are playing semantic games with "much" and "useful". Lower is always more efficient.

Your "add a ton of HP if you want faster response" has merit, even if it can't provide nearly as many btus as quickly as a larger buffer tank (the numbers are clear about that).

Posted By jonr on 04 May 2013 10:44 AM
That's right - you can send 95F water to the radiators while the buffer tank remains at only 80F.

So you are an Engineer without geo, radiant or building or building experience? That means you know enough to be dangerous here.
It shows in your comments which are good in theory but not good in practice, and some were quite a way off from making radiant in combination with geo work, and making it work efficient.
My questions are indeed "rhetorical", since they were meant to see how much you really know about radiant and geo, and your answers (or lack of such) were quite revealing.
Just take my last question: "So how do you run the zones at 95F and keep the buffer at 80F?"

I don't mind to exchange ideas here, but it is a fine line between theorizing and misleading people here because you lack the know how in the field.

Is there some part of "START ANOTHER TOPIC" that wasn't clear? But it really would be better if you figured it out - you have a well developed tendency to find a way to misinterpret just about anything and argue endlessly no matter how wrong you are (ref: 30F and 90F are always the exact right min/max loop design values). Come to think about it, if anyone else is interested, ask in the radiant forum and I'll respond. Docjenser - I just don't care.

One doesn't have to read much to find a lot of examples of installers, who should know both theory and practice, falling short on the former.

To Danreed76, the OP:


The following suggestions were made here, which were either experimental, theoretical or I would question wether some would work at all. They include:



1) ...chilled radiant floors combined with good air sealing (which ensures minimal latent loads even if it is humid outside) and ERV/DOAS system delivered dehumidified air is a viable (and probably superior) radiant cooling design..Caution - you do not want to open the windows with this design and concrete slabs.

2) If you use fans coils, you are eliminating the efficiency/delta-T advantages of large radiators. But solar gain, internal gains, aux heat and latent load should move the numbers somewhat closer together. Probably for the entire house.

3) With the weather in GA, I would expect 2x heating vs cooling to be a good starting point for sensible load. But solar gain, internal gains, aux heat and latent load should move the numbers somewhat closer together.

4) Consider not oversizing to 5 tons of heat pump for a 3.7 ton heat load (before aux heat, less after) and a less than 3 ton total cooling load.....Note: heatloss was 44KBTU/H, geo heatpumps are usually rated 9600/ton for heating

5) pressurize the house

6) use a 500 gallon buffer tank, and this will allow you to reduce the size of your heat pump

7) for off-peak, an underground cistern makes sense
8) you can send 95F water to the radiators while the buffer tank remains at only 80F

I would urge caution to follow any of those suggestions, from my experience they are either experimental or not beneficial to what you are trying to accomplish, namely to utilize radiant floors for your comfort (it also needs enough capacity for adequate response) and combine it with a geo system for efficiency, and ensure that you have a cooling system which can pull enough humidity out of your humid air in south of Atlanta climate, so you don't have condensation issues.

After what I've been able to glean from the thread and other offline research so far, here's what I'm thinking:


5.5 or 6-ton (two stage) WTW unit with 80 to 160 gallons of thermal storage (tonnage based on heating capacity of heat pumps at 65 degree EWT)
radiant floors throughout for heating only
Multiaqua or other type high wall mount hydronic fan for remote zones to avoid chases and ductwork (humidity control settings allow for variable fan speed to manage latent loads)
3.5 ton hydronic air handler for the first floor and basement with all ductwork in the basement (humidistat in conjunction with variable fan speed and water flow allow for management of latent loads)
ERV for outside air exchange.
DHW supplemented by desuperheater on GSHP. The jury is still out on the ground loops, I'm leaning toward open loop or horizontal.

I figured this would give me the capacity needed to maintain the house comfortably, buffer a bit for the latent load as I'm not sure how to calculate ACH-kids (defined as the continuous attempt to realize the life expectancy of exterior door hinges).

I like the concept of radiant cooling systems, but the more I read on them, the more I find that residentially they end up being failed experiments that nobody wants to take credit for afterwards. In certain environments, they appear to be excellent (arid climates, clean rooms, laboratory facilities, hospitals, etc). On a 95 degree afternoon in Georgia after a summer rainshower, however, I believe they would result in venomous conversations with the spousal unit.

That being said, I'd like to carry the conversation further regarding system architecture and controls. Best practices in setting up such a system, such as zone pumps, zone valves, controllers, mixing valves, etc.

Would you recommend using two seperate loops (one heating, one cooling) or just one supply and return loop with individual zone controls?

5.5 or 6-ton (two stage) WTW unit with 80 to 160 gallons of thermal storage (tonnage based on heating capacity of heat pumps at 65 degree EWT)

That's good precise writing. As you know, some people think of "nominal tons" which is a fairly meaningless number that isn't at all the same as tons. There isn't one number to convert nominal tons to btu/hr, it depends on the conditions. For example, perhaps 14,000 btu/hr of heating capacity per nominal ton for a HP using open loop in your area. But there are always exactly 12,000 btu/hr in 1 ton and this is the ton everyone should refer to. Better yet, just use btu/hr so that someone can't misconstrue it.

I agree with you on radiant cooling. Alone it can't work, with small ducted assist (DOAS) it might, if you had a good designer and well bounded conditions. But neither may apply to your case.

Two loops is more flexible. Maybe think about a zone for hot water vs just DSH.

At 50F entering water temp, a 5-ton hydron Heat Pump (Enertech) is actually rated at 66,200 (at 50F EWT) for heating and 62,700 (at 59 F EWT) for cooling. So with an open system running with 64F EWT you should have year around heating capacity around 70 KBTU/h and cooling capacity around 60 kbtu/h, much more than your loads account for. Take into account that you are loosing about 5 kbtu/h heating capacity when the DSH is running in the heating season. You can make this call when you have finalize your load numbers.
Now with a horizontal loopfield, your EWT might drop down to 30F during winter and go up to 90F during summer. Now your capacity drops to 50 KBTU/h (30F EWT) in heating and 50 KBTU/h (90F EWT) in cooling mode. Still well within the range of the load you had posted here.
http://www.hydronmodule.com/images/unit_literature/20D082-02NN.pdf
Other heat pump manufactures are similar. The problem is that if you make the jump from 5 to 6 ton nominal, I am not aware of a larger than 5 ton unit at the US market which qualifies for the tax credits. The only exception are the Waterfurnace 7 ton high temp and the 10 ton unit, both with the new plate heat exchangers, but that would be overshooting for you. None at Enertech, Climatemaster, Bosch etc
Giving the HP numbers above, you want to consider reducing your buffer tank to about 10gpm/ton as recommended by the manufactures. I think 80 - 160 is still too large, you don't store much, you simple add thermal mass in the radiant system which reduces you response time.
http://welserver.com/WEL0602/ 28 ton hydronic system, 2 x 80 gallons
http://welserver.com/WEL0664/ 7 tons, 80 gallon buffer tank
http://welserver.com/WEL0396/ 6 ton, 50 gallon buffer
http://welserver.com/WEL0383/ 6 ton, 50 gallon buffer

You can see that we moved away from zone pumps to a single pump on the load side (variable speed) and zone valves, as soon as a zone valve open the pump notices the pressure drop and revs up, keeping the pressure constant. No mixing valves, you want to keep the temps the heat pump is exposed as low as possible.
Unless in commercial applications with significantly different loads, using one loop should be fine. You can entertain using separate cooling tanks incase your spousal unit has a temperature tolerance of less than 1 degree F, heating to 72F, then start cooling at 72.5F.
But normally when you heat to 70F and start cooling around 76-78F, you don't really need it. You can have a manual switchover, flipping a switch, which makes the control design much easier.

That being said, I'd like to carry the conversation further regarding system architecture and controls. Best practices in setting up such a system, such as zone pumps, zone valves, controllers, mixing valves, etc.

Here and here is a starting design to consider. Of course the boiler is replaced by a HP. Note that it allows the buffer tank temp to be lower than the temp going to the radiators. Imagine that, a large buffer tank doesn't slow response time when you plumb it right. Plus you pick up a little extra efficiency because the coolest (hottest in summer) temps (from the radiators) are often going directly to the HP (ie, not getting diluted down by the buffer tank).

Thanks for the link, jonr. I hadn't found that website yet. I didn't realize that the buffer tank was plumbed as such. I guess for some reason I had it in my head that it would be more or less an indirect water heater that would be the "beginning" of the distribution circuit, and would be isolated from the heat pump by a coil, and an aquastat on the tank would define when the heat pump turns on/off. (I'm not saying it's right... it's just what I thought).

that being said, is stratification an issue in buffer tanks? should the system draw from the top and return to the bottom in heating mode, and vice versa for cooling mode, or did I just overcomplicate the buffer tank? (a great deal of my background is in hydraulics... in practice, tank temperatures can be tremendously different from top to bottom, defining where we place in-tank coolers, dip tubes, baffles and circulating pumps.)

Posted By jonr on 06 May 2013 09:42 PM

That being said, I'd like to carry the conversation further regarding system architecture and controls. Best practices in setting up such a system, such as zone pumps, zone valves, controllers, mixing valves, etc.

Here and here is a starting design to consider. Of course the boiler is replaced by a HP. Note that it allows the buffer tank temp to be lower than the temp going to the radiators. Imagine that, a large buffer tank doesn't slow response time when you plumb it right. Plus you pick up a little extra efficiency because the coolest (hottest in summer) temps (from the radiators) are often going directly to the HP (ie, not getting diluted down by the buffer tank).

http://www.radiantandhydronics.com/ext/resources/Plumbing/Home/Images/1107-portal-GF-Fix-lg-edit.jpg

I guess you confuse boiler applications with LOW temperature radiant in combination with GEO heatpumps.
You cannot simply replace a boiler with a geo heatpump, which works with a low delta T and a high flow rate.
A 5 ton heatpump, which is the point of discussion here, needs about a 10 gpm flow on the load side, puts out about 50 KBTU/h and has a temp delta of 10F.
Imagine now in your example what will happen if only one zone calls for heat, lets say danreed's master bedroom.
How are you manage to pump 10 gpm through one small zone? And even if you put in a huge pump in order to do this, a single small zone cannot take 50000 KBTU/H which the heatpump is producing.
So your delta T trough the zone would be minimal, maybe lets say 10 KBTU/H (which it does not even do at 95F supply), so your water coming back to the heatpump now is 93F, going out of it now at 103F. You know what that means? Severe SHORTCYCLING!

Other scenario: To ensure enough flow to the HP, P3 has to run pretty much all the time, now you are mixing the buffer tank you wanted to keep at 80F with the 95F degree water("That's right - you can send 95F water to the radiators while the buffer tank remains at only 80F") having less than 85F going to the zone (BTW, we have now 3 pumps running for 1 small zone). Now the water going back to the buffer tank mixes with colder return water from the zone and hotter water from the heatpump. It will do nothing but adding thermal mass to the system, decreasing response time. But it will not send 95F to the zones unless the buffer tank is at 95F as well.
Why do you think every single manufacturer has a simple circuit going from the heatpump to the buffer tank and back to the heatpump in their installation manual? And then the zones take whatever amount of BTUs the need out of the tank. Flow separation between zones and heatpump, both having different flow requirements!
Now imagine those scenarios when one of Danreed's 1 ton hydronic air handler turns on and can take only 3 gpm in A/C mode?

No matter what you do, you do not gain any efficiency because the supplying and returning water gets mixed with the buffer tank, and simply do not reduce response time. All it does is having many circulation pumps running, consuming electricity.
You have to ensure that the flow requirements of the heatpump (10 gpm) is satisfied all the time, no matter how many zones are running. If P3 is running all the time, that is assured with the design you are showing, and it will protect the heatpump. But why doing it when you don't gain anything since you are mixing the water anyway, since you have to mix the water to ensure flow rate?

should the system draw from the top and return to the bottom in heating mode, and vice versa for cooling mode,

As drawn is best for winter. I agree, reverse for summer.

Your fan coil units may require higher temps than you want to send to a radiant floor. Mixing isn't the most efficient way to do this. Also make sure that a large cold concrete slab doesn't ever cause the fan coils to emit cold air.

You should take Doc's advice on everything here HE knows what he is talking about.
I would add that you want to check temps and capacities for cooling on your hydronic air handler.

Posted By danreed76 on 06 May 2013 10:20 PM
Thanks for the link, jonr. I hadn't found that website yet. I didn't realize that the buffer tank was plumbed as such. I guess for some reason I had it in my head that it would be more or less an indirect water heater that would be the "beginning" of the distribution circuit, and would be isolated from the heat pump by a coil, and an aquastat on the tank would define when the heat pump turns on/off. (I'm not saying it's right... it's just what I thought).

that being said, is stratification an issue in buffer tanks? should the system draw from the top and return to the bottom in heating mode, and vice versa for cooling mode, or did I just overcomplicate the buffer tank? (a great deal of my background is in hydraulics... in practice, tank temperatures can be tremendously different from top to bottom, defining where we place in-tank coolers, dip tubes, baffles and circulating pumps.)

You don't need a coil to separate the buffer from the heatpump, every time you go through a heat exchanger you loose about 10% efficiency. I would recommend a digital boiler control for the tank, you can dial in the temp better, you can see if you end up with a 2 stage compressor or single stage which you need to control. Stratification is not an issue, you are pumping 10 gpm through the buffer tank when the heatpump is running, plus whatever zone is on. It will be mixing well.
I like to push in the bottom, should sediment formation ever occur, you don't pump it through the radiant system.
You plan to use radiant everywhere in the house, correct? If so you don't need to worry about the air handler temps, since they will only be on in cooling.

"Heating cycle:

Basement radiant slab
First floor radiant staple-up with diffuser plates
ERV installed in mechanical room to supply basement and first floor.
Second floor: hot water supplied to 2 each air handlers with ducted outside air exchange (1 for each 2nd floor wing as the great room vaulted ceiling splits the upstairs)"


From page 1. I am just trying to help with info on this long thread.