Need some help with an existing Radiant / Baseboard system

ScottFP

The globe valve would need to go after the pump like Jamie said correct? ... and as so i'm guessing that i need to install a globe valve on each of my secondary zones as a way to regulate the flow that each zone will use? ... should those globe valves also go directly after the pump, and the ball valves are used only to completely shut off the zone for service, etc.

I did the calculation for the head loss in the baseboard circuit again and came up with the same number. 180 degree water through 150' of 3/4 PEX at a flow of 1 GPM to equal approx 1' of head.

But were you saying that it would heat ok and the return would be hotter if i let the full 3.5 GPM from the primary loop flow through the baseboard zone and i won't have as much of a drop in temperature before the start of the next loop? Wouldn't that be a good thing?

Also 3.5 GPM in the primary loop is sufficient to supply 1 baseboard zone with 1 GPM, and 2 radiant zones with 1 GPM /Ea. correct?

Does it make a difference if the fast fill is connected to the air / sep. and exp. tank or would it make a difference if i left it where it is currently coming into the return just before the boiler?

Cliff Brady

Proportional mixing

I think the Pros here (I am not one of them) are trying to point you in the direction of a proportional reset piping scheme using mannually set 3-way mixing valves (TMVs are old school) on the lower temp radiant circuits like in figure 5 of this article:

http://www.pmengineer.com/CDA/ArticleInformation/features/BNP__Features__Item/0,2732,101268,00.html

Idealy, as a modification to Figure 5 you would also seperate your boiler loop from your parallel primary with outdoor reset controlled injection mixing.

BTW, your schematics are great. You have learned quite a lot.

Jamie_5

flow

Scott:

I'm going to be a little long-winded, but I hope it'll help clear things up.

Assuming the fintube baseboard can emit the heat to the air, you'll get the same *overall* temperature drop whether you are circulating 1 GPM through that circuit or 3.5 GPM. You have a 10,000 BTU/hr. demand on that circuit. Water has a specific heat of 1 and weighs about 8.33 lbs. per gallon. A BTU is the amount of heat it takes to raise 1 pound of water 1 degree F. Conversely, if a pound of water loses 1F, it has shed 1 BTU. So moving 60 gallons per hour (1GPM) allows you to transport 500 BTU/hr. per degree drop in temperature of the water. If you move 1GPM of water through the circuit, it will lose 10,000 BTU/hr. and drop 20F in temperature. So, if that water were at 180F, it comes back at 160F. 1 gallon of 160F water mixed with 2.5 gallons of 180 water equals 3.5 gallons of 174.3 water. So the 1GPM that went through the baseboard circuit mixes with the 2.5 GPM that didn't to give a flow to the next circuit of 3.5GPM of 174.3F water. If you move all 3.5GPM through the baseboard circuit, it will drop 5.7F to give off 10,000 BTU/hr., giving the primary loop the same 3.5 gallons of 174.3 water. So the next loop will see the same temperature water either way.

Remember your original problem, namely very low return temps from the radiant zones. You correctly, I think, diagnosed this as a result of low flow. That is because the heat is moving from hot to cold as much as it can until the temperature difference between the tubing in the joist bay and the joist bay itself is zero. (The poster who said the returns couldn't be 1/4 the temperature of the supplies was undoubtedly correct. As he pointed out, that would mean that the return temps were 40 to 45F, which would imply that the water and tubing were getting colder than the air in which they resided, which is impossible.)

The reason for the globe valve is not to control the water temperature so much as to control the velocity of the water in that circuit. The accepted recommendation is to keep the velocity between 2 and 4 feet per second. Significantly higher, and you can get flow noise and erosion of the piping, as well as excessive resistance (head). Significantly lower, and air does not always stay entrained, leading to possible air pockets. 3/4" tubing can move about 4GPM at a velocity of less than 4 ft. per second and without excessive resistance. So you want to limit your flow to that. Hence the recommendation for the globe valve to restrict the flow in that circuit.

The other two circuits are very close to the flow you want in each using the 007 pumps, so I don't think you'll need valves to adjust the flow on those. Each radiant zone has 4 parallel circuits requiring, you say, 1GPM each, for a total flow of 4GPM. (Calculating as above, you should see a 6.6F drop in water temperature across the whole zone, which is safely within the 10F range often used a design baseline for radiant floors, though you could probably use a greater drop in a joist heating application without experiencing "heat striping," i.e. noticeable temperature differences across the floor surface.) Because the loops are in parallel, you design around the most restrictive loop. You need to move 1GPM around the most restrictive loop at about 9 ft. of head. (According to Wirsbo, 1GPM of 160F water encounters/produces 0.03237 feet of head per foot of tubing. Multiplied by 270, your longest loop, gives 8.74 ft. of head. There will be some additional head imposed by the short runs of copper tubing.) The 007 will move about 4GPM at that head, just what you want in each radiant zone. The flow across each loop will not be exactly 1GPM, but because the loops are close to the same length, it will be close enough for government work. So on those zones a globe valve will just be a further restriction to flow and unnecessary.

As to your primary loop requirements, you need 4GPM in each of two radiant zones and 1GPM in the baseboard zone. That's equal to 9GPM total. But your primary loop doesn't need to flow that much because it only needs to move 32,500 BTU/hr. At 3.5GPM, the water will only drop 18.6F in temperature to give off that amount of heat. So 3.5GPM through the primary loop is sufficient. You can get away with this because the water temperature you need in each radiant zone is *lower* than the supply water temperature in the primary loop. So the radiant zones will not need all 3.5GPM of primary loop water to flow 4GPM of 160F degree water. The first radiant zone will see supply water of 174F if the baseboard zone is running and 180F degree otherwise. The temperature drop across this zone at 4GPM will be 6.6F, equal to 13,250 BTU/hr. divided by 500 (the number of BTU/hr. 1GPM of water can move, divided by 4, the number of gallons you're actually moving through the zone). Your supply temperature is supposed to be 160F, so your return will be 153.4. You need about 1.3GPM of 174F water to mix with 2.7GPM of return water at 153F to provide 4GPM of 160F water. The first radiant zone will therefore return 1.3GPM of 153F water to mix with the other 2.2GPM of 174F water that moved right by to produce a supply of 3.5GPM of about 166F to the next zone in the primary loop. To get an output of 10,000 BTU/hr. at 4GPM in that zone, the water temperature will drop 5F, for a return temperature of 155F if the supply is 160F. You will need just a little less than 2GPM of 166F water to mix with just a bit more than 2GPM of 155F return water to produce 4GPM of 160F water. The primary loop has sufficient flow of sufficiently hot water to do that. The slightly less than 2GPM of 155F return water from that last zone will mix with the 1.5GPM of 166F water to produce 3.5GPM of water at just over 160F. That is exactly what we expected to see from 3.5GPM of 180F water giving off 32,250 BTU/hr. because 32,500 divided by 500 divided by 3.5 is 18.6, which is the temperature drop we need to move that many BTU/hr. at a flow rate of 3.5GPM, and 180 minus 18.6 is 161.4F.

Everything I've just written is why I think your series loop primary/secondary system should work if you pipe it as you pictured it in your previous post.

The reason for placing the fast feed at the expansion tank is because the expansion tank is the one place in the primary loop where no pressure change is induced by the primary pump turning on or off. The pump produces a differential in pressure between its inlet and outlet when it runs. With the pump off, the pressure is equal around the whole circuit. But when the pump comes on, it produces that pressure difference it is designed to make. For reasons explained well by Dan Holohan in his book "Pumping Away," the system pressure can't drop at the expansion tank. If the pump is immediately after the expansion tank, the pump adds pressure to the system at its outlet, which decreases as the water moves around the system, due to the resistance to flow imposed by the piping. In your case, there's not much pressure drop around the primary circuit except for the water heater, so it probably won't make much difference to place the auto feeder where you have it. But if you really want the feeder to maintain the system pressure your expansion tank is attempting to maintain, you put the auto feeder inlet at the expansion tank because the pressure there is always what you set the expansion tank to.

I hope this helps. Please let me know how you make out.

ScottFP

flow meter

Does anyone know of a place that i can purchase some fairly inexpensive flow meters?

That would make things a lot easier with balancing everything out after i get the necessary repairs made.

Thanks,
Scott

Jamie_5

flow meters

You shouldn't need them. The thermostatic mixing valves will take care of getting the right temperature water to the radiant zones. The over-all flow in those zones will be good enough if the numbers are right based on the pump size. The only zone you may be interested in limiting is the baseboard zone, and then just enough to prevent excessive flow. A thermometer on the return side of that zone will tell you what flow you have. If you want to know that the radiant zones have sufficient flow, you can place thermometers on the returns of those zones, on the zone side of the mixing valves. The difference in supply and return temperatures will tell you how much flow you have.

ScottFP

Check Valves

As per the diagram above ... The check valves on the two radiant zones, do they need to be placed to the left or right of the mixing valves? or does it make a difference?

ScottFP

Some repairs made

I have made all the changes in the new diagram today, except the primary pump swap. It is on order and will be here monday. Hopefully everything will be working well then.

Thank you all for the input and help.

Scott

Jamie_5

changes

I, for one, am curious to know what happens. Please report back.

I suggest no setback on the radiant zones, especially not the carpeted one.

Good luck.

Tom R.

Watts mixing valves

Watts MMV and 1170 are recommended for radiant heating. The 70A is for domestic hot water and flows only 10 GPM. The TACO 007 flows over 20 GPM. This could explain the large temperature difference. Those swing check valves are also a big flow reducer. Use a weighted flow control valve instead.

ScottFP

Ghost Flow Question

Today i changed the primary pump to a Taco 009, and everything seems to be working. however when one zone is on the other two have hot water flowing through them also. I have a check valve on both the supply and return of the 2 radiant loops, and one on the supply of the baseboard loop. However they are swing check valves, and after some reading on primary / secondary design, this type is not correct.

So my question(s):

1. can i use a spring check valve in place of these swing check valves to fix the unwanted flow?

2. Do i need to put a check valve on the return of the baseboard loop?

3. Do the check valves need to go on the secondary loop side or primary loop side of the mixing valve or does it make a difference? (As per the diagrams, to the left or right of the mixing valve)


Thanks,

Scott

Jamie_5

check valves

Congratulations on the progress you've made. So you're getting heat in the radiant zones now and the return temperatures from those zones are reasonable? If so, it must feel good to have accomplished that.

Siegenthaler's book suggests using either flowchecks or spring checks on both the supply and return of the secondary loops. I don't know that it makes any difference, but I would place the checks outside of the mixing valves, i.e. to the right of them in your drawing. And yes, I would use checks on the baseboard loop. But maybe someone who has actual experience will suggest differently.

Having said all that, are you sure it's thermosiphoning you're seeing? That is, are you sure that the other pumps are only coming on when they are supposed to?

ScottFP

The temperatures are just about exactly what you calculated in one of the previous posts, so it seems like everything there is working okay. I will change my existing swing check valves to spring check valves, and place one on the baseboard zones return. The pumps are only on when the thermostat for that zone requires heat, all that is taken care of automatically through the Taco relay, and i checked it out; they all seem to be running properly.

Jerry Durnin

Thermostats

My hot water heat seems to warm the house well but the house seems to get cold before the heat comes on. I was told that my old honewell mercury thermostat has an adjustment on it. I found a place where the adjusment seems to be but I dont know how to adjust it.

ScottFP

yet another problem

The gas usage which was high before, is now almost double what it was. The system is using approx. 500 Cubic feet of natural gas to supply at most 35000 BTUs on a 40 degree day. So i placed a call to Takagi and talked to one of their techs today. He took all the information about the system (btu's, flow, etc) and calculated that at most the system should be using approx 250 cubic feet of natural gas (i'm pretty sure he was talking on a 0 degree day) but anyway .. he said that its possible that the gas line was not installed properly and that there is too much pressure going into the water heater. Which could have possibly screwed up one of the 5 valves and the modulation and the seals, etc... So... I will check to see if the he is correct and the pressure is to high. Hopefully this will be the last kink in the repairs.
Does all this sound feasable as a possible solution to the gas usage problem?

Jamie_5

gas usage

What I know about gas trains would fit on the head of a pin with room to spare. So I don't know what the effects of the gas line pressure would be. I still think of the two potential problems I mentioned a couple of weeks ago: the heat in the joist bay zones may be going where you don't want it, i.e. to the great outdoors if the ends of the bays aren't effectively insulated, and/or the water heater is not an efficient way to produce small temperature gains.

I don't know what the insulation is like at the bay ends, but if it is not good, increasing the flow in those zones will have made the gas consumption problem worse by increasing the temperature difference between joist bays and the outdoors. Hence the need for insulation.

As to the second issue, I looked at the installation manual for the water heater, and it appears that it modulates the burner in response to flow as much as to temperature. As a domestic water heater, it is designed to produce a large temperature rise in the water, on the order of 70F, and it's being asked to produce only a 20F rise. My guess, and it is only a guess, is that the unit is short cycling like crazy because the relatively high flow of 3.5GPM (high for a water heater) is ramping the burner up to high but the limit of 180F is being quickly met because the heater is operating at its maximmum 140,000 BTUH input rate when you require only 16,000 or so BTUH.

This is very likely why some of the professionals who posted suggested a buffer tank. But I think that even with a tank, there will still be a control problem. To keep the heater from cycling, you would need to allow a wide differential on the water temperature in the buffer and keep the flow through the water heater low when the buffer called for heat. I don't know how wide a differential you can have on the tank and still have a comfortable home, since the wider the differential, the lower the water temperature to the heating system when the buffer temperature is approaching the low limit of its aquastat. The lower water temperature would reduce the output of the heating system, and given the relatively slow response time of the joist heating zones, could leave those zones constantly "behind the curve" in terms of keeping the house warm.

It might simply be better to have an appropriately sized boiler, which would be designed to produce a smaller temperature rise at a higher flow rate.

ScottFP

RE: gas usage

The joist bays have a doubled section of the r19 insulation at the end of each joist bay. As for the boiler, it doesn't short cycle it runs as i would expect it should, and the guys at Takagi seemed to think the supply of 180 and the return of 160 sounded normal for a space heating setup with a primary secondary piping design.

Jamie_5

RE: gas usage

Scott:

All right, I'm out of ideas. Maybe try starting a new thread asking about the gas train?

Just to be sure the gas usage is way out line, let's do another *very* rough, back of the envelope calculation. 500 cubic feet of gas should contain approximately 500,000 BTUs of energy, but you didn't say over what period of time that was used. Conduction and infiltration losses are directly proportional to the temperature difference between the outside and inside. If the average outside temperature was 40F, that's about half the design temperature difference and thus approximately half the heat loss at design temperature, which was roughly 37,000 BTUH. So you should be losing about 18,500 BTU per hour at 40F. If the system is 80% efficient, which is probably very optimistic, it would take 23,125 BTUs per hour to meet the need, or 23.125 cubic feet of gas. So it should take about 22 hours to use 500 cubic feet of gas if the temperature remained constant. How long did it take to use the 500 cubic feet you consumed?

I'm sorry this is such a trying project. I hope you figure it out.

ScottFP

RE; gas usage

about 500 cubic feet were used in 24 hours, if that is correct, then the price of heating this home with natural gas seems Extremely expensive, This is the first home we have ever lived in that utilized natural gas, oil and coal are the only things i've had personal experience with. At the current price of 100 cubic feet of gas being somewhere around $1.75, The monthly heating cost especially if the temperature were around 0 degrees for a month would be insane. 10.5 CCF (or 1050 cubic feet) in 24 hours; $550 or so? Does this seem normal for a moderate sized decently insulated home built around '86?

ScottFP

Re Re: gas usage

Also the system is setup to output 35000 btu / hr when all 3 zones are running, so if the system ran for 24 hours / day then it would consume about 500,000 BTU's, but it only runs maybe a half an hour every 2 hours. this confuses me ) ...

Jamie_5

RE: gas usage

First caveat. My calculations were very rough. For instance, they assumed that the temperature outside was a constant temperature. They also assumed the heat loss estimate was accurate and they do not account for variables such as occupancy, lights, and load changes with seasons caused by cooler ground below the basement, cooler external mass of the building, etc.

Second, the design temperature used to calculate the heat loss is the temperature seen 1 to 2.5% of the time during the year, depending on the approach of whoever did the heat loss estimate. You should not see a solid month of 7F weather in your location, so you will not see the use you calculate. (As reflected in the heat loss you posted, 7F is the design temperature for State College PA. I believe that it is warmer than that 97.5% of the time there.) So your fuel usage should not be as bad as you calculate.

Third, gas prices are high right now.

Fourth, the correlation between the heat load estimate and the run times will only be perfect if the heat loss estimate were perfect and the system design was perfectly matched to the design load. Both are very unlikely. So the run time of the system won't perfectly match the estimated heat loss.

Fifth, the BTU's required are independent of the fuel choice or the heating medium. If the house loses 37,000 BTU per hour when the outside is 7F, the heating system has to replace that much heat, which it does by burning gas (in this case) at some combustion efficiency between 80% and 90%, hopefully. If the system could achieve a perfect steady state, which it won't, it would require 37,500 BTUs x 24 hours divided by the efficiency on a day where it was 7F all day. Assuming 85% efficiency, that's over a million BTUs on a design day. Water or air, oil or gas, it would still take that much energy to heat the house.

Sixth, it's not clear to me that you accounted for other devices that use gas when you metered the usage.

To me, it sounds like things are relatively normal now. If you want to save money on heating costs, work on insulating the house better.

The important question now is, "Is the house comfortable?"

Tony_23

Heating hours

are a conversion from degree-days. I believe Stae College, PA would be in the 2000 hour range.