Buffer Tank Size

Housedoc

Using the Uponor buffer tank sizing recommendations, I came up with a 50 gallon requirement on my geothermal radiant project. However, the Water Furnace installation guide says to size the buffer tank at a minimum of 2 gallons per 1000 btuh capacity. This means the tanks should be 72 gallons. What gives? Why is there a difference in the two publications? is there a disadvantage to oversizing buffer tanks other than jacket losses from a larger vessel? Which size is right?

furnacefigher15

different maufacture

The purpose of a buffer tank is to stabilize load fluctuations.



Go with what waterfurnace wants.



The 50 will probably be fine, but may cause shorter cycling if the load fluctuates heavily.

NRT_Rob

I have no idea

how Uponor sizes buffer tanks. Typically you size for a minimum runtime you are trying to achieve.



with Geothermal, you can't use a large temperature rise, so you typically need more volume to store the same number of BTUs. I think normally we are sizing Geo buffer tanks at a 15 degree rise or so.

Housedoc

Me neither

I am not sure they know. The 50 gallon gives me a 14 minute cycle time using a 10 degree TD. I could use an 80 if I need to. It stretches the run time to 23 minutes.

NRT_Rob

really?

you only have an 18kBTU output? 1.5 ton unit?



Otherwise, something is wrong with your math.



50 gal x 8.4lb/gal x 10dt= 4200BTUs stored.



4200/14 minute = 300 BTU/minute rate of charge.



300 x 60 minutes = 18kBTU/hr output rating.

Housedoc

3 ton

No, its a 3 ton Water Furnace Synergy III, 2 stage. Although hydronics always goes to high.

The load is 17,000 btuh. Uponor's calculation shows that you use the heat loss instead of the unit's capacity. I guess that makes sense since you only heat the full 50 gallons at start up. From then on, it's only maintaining 10 to 12 degrees used in the floors. So on a design day, it will run about 14 minutes. The rest of the time it cycles for shorter periods.



I guess I could use a wider differential for the tank control. But I don't want to miss my target temp for the floors.



I am inclined to bump the tank to 80.

NRT_Rob

not correct

you do not use the heat loss: you have to buffer during part load conditions where loss could equal very nearly zero. You use the output of the unit at zero, or nearly zero load, unless you can be SURE of the attached load. for instance if you are attached for a forced air hydro system you know extraction will be X btu's/hr and if you had a minimum fan time you could be sure of that component of your runtime.

Housedoc

Uponor Calculation

Rob,

The calculation I have from Uponor shows -

Vmin = Cycle Time x ( HP Capacity - Load BTUH) / DT x 500



Vmin - minimum buffer tank volume.



Assuming a cycle time of 10 minutes.

10x (38,000 - 17,000) / 5000 = 42 gallons



A 50 gallon gives me about 12 minutes.

An 80 gallon gives me about 19 minutes.

Of course this is at design temperatures.



Which one would you use?

NRT_Rob

unless you have a guaranteed load

like a forced-time on air handler, use load equals zero for this calculation.



that is why your calcs are not matching expectation.

Housedoc

Dont get it

Rob, I am not getting what you are saying. The load is guaranteed as long as its cold outside. It's 17,000 btuh on a 20 degrees F day. The calculation indicates the btu's in a 50 gallon tank will supply the design btu's for about 12 minutes. The 80 gallon will last another 6 to 7 minutes. The unit energizes and replaces the missing heat in about 30 minutes. Am I looking at this wrong?

Gordan

Yep, you are looking at it wrong

The purpose of the buffer is not to provide heat or to extend time between cycles during design load conditions, it is primarily to ensure that the duration of a cycle during minimal load conditions is above some minimum required for efficient and reliable operation. It does this by absorbing the difference between the smallest expected shoulder-season load and the output of the heat pump (which is sized to the design load.)



So... size the buffer so that it takes the desired amount of time to raise its temperature by the appropriate temperature differential with input of (heat pump output - minimal load).

Housedoc

Run Time

So, if the tank is 80 gallon and the machine is 36,000 Btuh output with the aquastat set on 120 with 10 degree differential and the load is 17,000 Btuh, how long will the unit run to satisfy the cut out temp?



It looks to me like approximately 11-12 minutes with no load and 19 minutes on a design day.



Is that correct?

NRT_Rob

again

yes, with no load you are at 11 minutes. 21 minutes at design temp.



if you are really running 120, I hope you're using a reset control. waste of good geo efficiency 90% of the year to run at such a high temp.

Housedoc

Outdoor reset control

I have not gotten that far yet. I am using the 80 gallon tank. But have not determined how to control the tank temp. With such a small load (17,000 btuh), is an outdoor reset controller worth the added complexity? What product would you use?

Housedoc

Taco delta T

I am using the Taco delta T pump on the radiant loop set to maintain 10 degree F temperature difference between supply-return temperature. Considering that, would an outdoor reset controller provide any measurable savings?

NRT_Rob

reset control

vs 120 fixed temp geo? you can easily add a an average of a COP for the season with reset. should be a 1 year payback or less on most systems, and it's not very complicated to add. if you can save an aquastat payback is exceedingly fast.



I will never understand why anyone spends the money on geo without reset. it's a small dollar add to really realize the efficiencies that geo promises.

Housedoc

Outdoor reset control

Okay Rob,

You convinced me. I ordered it today.



Thanks for sharing your expertise.