Sizing circulator for mixed hydronic gravity and new piping

Steamhead (in transit)

for the kind words!

Dan, your best bet may be to use two circulators, which would also let you run them as separate zones. The Grundfos is about right for the old gravity zone. Not sure what to tell you about the PEX zone since I don't know how it's piped, but I'm sure its flow and head characteristics are much different from the old gravity zone. Specifics?

<A HREF="http://www.heatinghelp.com/getListed.cfm?id=367&Step=30">To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"</A>

Rodney Summers

We built a new addition to our 1914 house and at the same time we replaced the original cast-iron boiler. The system had been converted from gravity to forced circulation in the past. The new system is now a combination of the existing gravity pipes feeding the original rads and new cast iron rads fed via flexible IPEX piping.

The new boiler is a Viessmann Vitodens 200 WB2 8-32. We have about 380 EDR of rads on gravity pipes and 400 EDR of new rads fed by IPEX. The heat loss for the house is 70 MBH.

The circulator that was installed is a Grundfos UP15-42F. The reason I believe the circulator is undersized is that I had to restrict flow considerably on most rads in order to get hot water to reach some rads of the second floor. Would moving to the UPS15-58F solve the problem or would we need to move to something much stronger like a 26-64F? How can I properly determine the size of the circulator pump that should be installed?

Glenn Harrison_2

Please check out this article...

in the Hot Tech Topics section of Heating Help .com. It should give you some insight as to what pump should be on your converted system.

Sizing Circulators for Hot Water Heating Systems by Frank (Steamhead) Wilsey

Plumb Bob

If you restricted the downstairs flow to get water upstairs, that's great, now you have sort-of-balanced the system.

A bigger pump doesn't help. Take a simplified case with only two parallel loops. The water can take either the "short" path (short length, large pipe ID, or both) or a "long" path. If these paths are significantly different, it doesn't matter how powerful the pump is, water will always take the short path.

You can solve the problem two ways. One is to make the resistance of the two paths the same, by restricting the short path, as you say you have done. The other is to use separate pumps for the two paths (they could in many cases be the same model pump, but you still need two distinct pumps).

Either of these methods will work, but even a humongous pump will not work like of two separate pumps. That's hydrodynamics for you.

viccipe

Gravity pipes now pumping

Normally large pipes are used for gravity back in the day.
So u button up by adding expansion tank and pump etc.
The pump normally good for high volume flow not pupmping pressure. Your new Kitec loop should work fine with regular up-1542 brute.

2 pumps and balancing should be good to go!

Rodney Summers

Yes, I had seen Frank's good article about sizing pumps for gravity conversion, but I did not feel I could use it as-is since my system is a hybrid. I agree the 15-42F would have been adequate to heat the gravity pipes.

One thing I should clear up is the number of zones. The gravity system had two large pipes leaving the boiler (one for the front of the house and one for the back). All the new rads were added as a single zone each (ie the IPEX pipe leaving the boiler serves only one rad). I thus have 9 pipes leaving the boiler, the two large gravity pipes and the 7 IPEX pipes. All pipes connect to the same Viessmann low-loss header and I am able to restrict the flow of each pipe individually at the boiler.

My concern about restricting the rads on the gravity pipes is that does it not mean that the rads are only emitting a fraction of the radiation they are capable of? It would mean having to run the Vitodens at a higher temperature to radiate the same amout of btu/h they were designed for.

If I understood some posters it sounds like the sizing of the pump has more to do with the expected total load based on the heat loss of the house rather the number of rads and size that need to be heated? Is it simply to say that 70 MBH should result in 7pgm for a deltaT of 20?

Brad White_31

Chiming in, Dan-

Firstly, congratulations on the Vitodens. And you are right in wanting to resist raising the temperature. Always a noble pursuit.

Question/Suggestion: It does not seem that you used the low-loss header to decouple the two distinct pressure zones. Thinking primary-secondary perhaps with the radiant right off the boiler and the rads off the header might work better for you. The Vito literature has a number of diagrams; were they followed?

I agree with Steamhead on use of separate circulators. Often if not always I will design radiant/PEX curcuits with a separate manifold and circulator because they do have higher pressure drops. In your case at least you can throttle back the circuit flows, radiation and PEX circuits. The other comments in this regard support this.

And as J. Cricket states, a larger pump is not the answer. It just means more throttling. Flooring the accelerator and riding the brake... Gentle persuasion and balancing are key. A separate circulator for each pressure zone, appropriately sized for flow and pressure will serve you best, IMHO.

As far as limiting flow to radiators affecting output: I would not worry about it. The real variable that makes the output linear to input is temperature.

Output capacity relative to temperature is a function of the difference between average water temperature within the radiator to room temperature.

Output capacity relative to flow volume is decidedly not linear especially at low flows.

As far as flow volume, you can reduce flow by 50 percent and still get 80-90 percent of capacity. The heating capacity reduction from flow adjustment is a function of reducing the average water temperature within the radiator.

If you were running 140 degree supply water and a 20 degree drop (120 F return) your average water temperature is 130 F obviously. 60 degrees warmer than a 70 degree room.

If you were to cut flow in half you would still have 140F going in but the average water temperature would be 120 F., a 10-degree difference and still 50 degrees warmer than the same 70 degree room. Not much reduction you can see.

Looking at this another way, you can reduce the flow to 10 percent of a respectable starting flow and still get about 50 percent of output. At the high end you can quadruple the flow from the same starting point and only get another 10 percent or so additional.
Hardly linear, it is an accellerating curve toward the low flow side and practically flat on the high flow side. But if you play with primary temperature, control is yours.

(Do you get the idea that America is over-pumping?? )

Your question/postulate on pump size is essentially correct. Flow rate wants to be matched to heat loss in a hot water system. And one gpm flow rate at 20 degrees delta-T will deliver 10,000 BTU's per hour (within fractions).


My $0.02,

Brad

Paul Pollets

conversion

The 8/32 Vitodens will require the Viessmann Low Loss Header to accomodate flow rates. A 26-64 Grundfos pump or a 26-99 pump should work. The 15-42 is too small.

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Mike T., Swampeast MO

You have an interesting problem...

"Home run" piping (what you have for the new radiators) does not a "zone" make. A true "zone" has some form of control individual control for the radiator(s) served. Is there a wall thermostat controlled valve for each of the new radiators? Or perhaps thermostatic radiator valves (TRVs)? What about the old gravity portion?

The forces that enable gravity circulation are very slight--thus the piping has to be huge so that there is exceptionally little resistance to flow. BUT, these forces are affected greatly by elevation. While the absolute difference between flow ability to a ground floor and a 2nd floor radiator is tiny, the relative difference is huge (nearly double). Thus to achieve balance higher radiators had intentionally more restrictive piping than lower. Again, the absolute differences are tiny, but the relative differences are significant.

While it's very easy to think that that pipes in a gravity system are huge because they move a lot of water, this is not the case. They are huge to keep velocity very low--very low velocity means very low restriction to flow. The goal of gravity system design is equal velocity throughout the system regardless of elevation above the boiler. Low radiators have larger piping than upper radiators of the same size. Branches to low radiators are often taken off of the top of the main (where the water is hottest) while upper radiators are often taken off the side of the main where the water is cooler. Yes, the velocity in the mains is low enough and the pipes large enough that such stratification occurs in the piping.

When converted to forced flow, the gravity rules no longer apply. Since the system is closed and pressurized as part of conversion, elevation no longer has any meaning regarding flow restriction. So, you now have a system where some radiators have two or more times the relative restriction of others. This is BAD NEWS for a circulating pump as it ONLY cares about restriction to flow. Say the gravity system had an original maximum flow of 4 gpm. If you put in a circulating pump designed for 4 gpm at zero head loss you will not get even circulation as it will greatly favor the low, close radiators with the least restrictive flow path. The simple solution is to use a circulating pump that moves MUCH more water. By moving MUCH more water flow restriction in the system as a whole increases making the tiny absolute differences in the gravity piping virtually meaningless. Velocity is no longer equal in the system. This would be a disaster if only a small amount of water were moving, but again, you're moving MUCH more water than the original design.

The affect of this much higher flow on the output of the radiators is complicated by the change that has also occurred in the boiler control.

This has already been a very long (yet still incomplete) answer to a seeming simple question, but restricting flow to the gravity portion does not mean that "the rads are only emitting a fraction of the radiation they are capable of". With enough restriction however it however possible to unbalance the flow terribly.

Your balance problem is most likely one of flow restriction. The new radiators (even with home-run piping) likely have significantly more restriction to flow than ANYTHING in the gravity system.

If you've throttled flow to the gravity portion to increase its' restriction so that you get flow through the new radiators you may not have enough flow in the gravity portion to achieve balance, e.g. the 2nd floor radiators of the gravity portion no longer heat well. If this is the case try greatly closing the hand valves on the lower floor radiators--you might get lucky.

If the 2nd floor rads that don't heat well are new ones with home-run piping then a higher capacity circulator might help. I can though nearly guarantee that you'll still have a tough balancing act between the new radiators and the gravity portion.

Separate circulators for the gravity portion and the new portion would probably work if piped correctly. I'd highly suggest contacting Viessmann for their piping recommendations.

Thermostatic radiator valves (TRVs) on all of the gravity system rads (at a minimum) would work as they provide their own variable restriction to flow and achieve their own balance.

Sorry for the length. If you reply with answers to my questions, we can try to help further with simple solutions.

A picture of the near-boiler supply piping would probably help as well.

Dan Miller_2

I will provide a picture of the near boiler piping when I have returned home.

The boiler supply goes into the low loss header. The pump is installed on the supply coming out of the header. The pipe then branches off to supply the 7 PEX pipes and the two gravity pipes. The returns all consolidate into a single pipe that feeds into the return of the low loss header.

The vitodens also supplies a Vitocell DHW indirect tank using the internal diverter of the boiler.

One of the rads that is not getting as much hot water is connected to the gravity pipes but it would have more head than the others because the radiator was relocated and piping with a few elbows was needed. The other rad is a "home run" PEX piping but it happens to be the rad that is the farthest away from the boiler so it would also have more head than the others.

If I can achieve a balance by restricting the branches with less head, should I still look into having a separate pump/zone to supply the two gravity pipes?