Welcome! Here are the website rules, as well as some tips for using this forum.

# NOT pumping away

Posts: 6
I am servicing a boiler standard single loop with 4 zone valves.  Here is what I must fix so far:

(1) pump is 6 inches from the output of boiler pumping toward the pressure tank that is about 2 ft downstream.   (2) feed and pressure reg valve 12 lb feeds into the  the output line of the boiler about 20 ft downstream from the pressure tank(set at 12 lbs.)

Question: could either of the above conditions be the cause of an 8 lb pressure rise in the boiler as soon as I get a call for heat and the heater fires for about 1 minute?  Pressure of warm boiler is about 12-15 lbs.  right after the pump kicks in and the boiler fires, the pressure rises to about 23 lbs.  (another 4-5 lbs and my bleed valve will start leaking)

Thanks,

Taggs

• Member Posts: 15,419
Sounds like

the tank is waterlogged- check that first.
All Steamed Up, Inc.
Towson, MD, USA
Steam, Vapor & Hot-Water Heating Specialists
Oil & Gas Burner Service
Consulting
• Posts: 6
Re: Water logged

Am climbing under the sub structure tomorrow to get an accurate piping diagram.  Have to get a equivalent length of my zones and check my gpm and pump curve.  I was thinking that I am not moving the water, also.  Thanks for the reply.  taggs
• Member Posts: 4,469
Re: Moving water

How many zones were you running while testing?
• Posts: 6
Re: Moving Water

• Member Posts: 7,356

Each loop has a head loss at some stated flow.  Add all the flows together and you have a system flow that the pump needs to provide.  That number is used to calculate the head loss through any shared piping.  The three parallel branches are summed in the same manner as parallel electrical resistances:  Add the reciprocals and take the reciprocal of the sum.
• Member Posts: 4,469
?

Add them together? Not the longest loop?
• Member Posts: 7,356
edited August 2012
parallel hydraulic resistance

Worst case for longest loop (even that assumes the loops are the same diameter) will get you close, and would apply at the single loop flow rate if all other loops were valved off.  The actual head presented to the pump when all loops are at full flow will follow the parallel resistance formula.  Note that full flow might be different for each loop depending not only on the length but also any balancing valves, pipe size changes, etc.  Floor loops of PEX make for relatively simple arithmetic.  Complex piping of different sizes with multiple branches will eventually reduce to a simple series-parallel diagram.

Edit: I'm seeing some use of different math for hydraulic resistance, which is even better for this than feet of head since it properly accounts for flow. http://goo.gl/fU0tv explains it far better than I can.
• Member Posts: 4,469
According to Taco

STEP 4: ESTABLISH THE HEAD LOSS OF THE PIPING SYSTEM AT THE TARGET FLOW

RATE

In a system with multiple zones, the head loss needs to be calculated for each parallel loop.The loop with the highest

head loss will be used in figuring the total equivalent length in Example 4.

Am I not understanding something?
• Member Posts: 9,514
edited August 2012

In parallel piping head will always find equalization through out the system loops. Flow will vary depending on the head of the loops, and how balancing is done. Pump size should still be calculated using the highest head loop including equivalent length supply, and return piping.

If people are adding heads of loop lengths together there are going to be some way oversized circulators out there.

If you want to check system head use a couple of 30 psi pressure gauges one tapped into the inlet side, and one on the outlet side of the circulator. The psi difference while the system is running multiplied by 2.31 will equal the system head.
• Member Posts: 9,514
Zoning with valves or circs

Zone valves will use less electricity than zoning with circs.

Staple up especially with out plates will use a need a higher water temp.
• Posts: 6

Still working on this installation (re-designing) Here is what I have found:

Longest loop 3/4in copper 277ft including fittings, but not zone valve (couldn't find value) main drop 1.25" copper 117 ft including fittings

Figured at 550 BTU per foot of fin coils, I have ~39K worth of BTU output into the house. Boiler output is 110,700 BTU.

Formula I used: HL=k X L X flow in GPM to the 1.75 power (4 GPM required for needed Ft/sec rate in 3/4in copper piping.

HL = .003 X 394 X 11.3 = 13.35

How do I figure my target GPM? Do I use the boiler size in BTU output divided by (450 "50% glycol" times my delta T of 20 OR Do I use my fin radiator lengths and figure the max BTU they can transfer plus my hot water heater sidearm BTU transfer rates?

I went with the Boiler size... let me know if this is wrong because I am guessing at this point. 110700 divided by (450 X 20) = 12.3 target GPM

Looking at a target of 12.3 and a head of 13.35, I don't think the existing taco 007 can do the job. Question is where do I go from here?

I'm thinking the taco 0012, but that is a monster pump! What happens when only one zone is open with a 8 lb head, does the 0012 push 35GPM though this loop at 1000 ft/sec? Seems it would be a bit hard for the fin coils to reject heat from the water moving that fast.

Any and all direction from here would be greatly appreciated.

Thanks
• Member Posts: 7,356
first things first

> Figured at 550 BTU per foot of fin coils, I have ~39K worth of BTU output into the house.

550 BTU/ft assumes 180F water, which will work, but If you can lower that temp (reducing the BTUs/ft) you will increase the system's efficiency.  Doing so may not be practical in this case.

> How do I figure my target GPM? ... Do I use my

fin radiator lengths and figure the max BTU they can transfer

Correct, though you also want to be sure that 39,000 BTU/hr is enough to heat the space.  Do you have a heat loss calc?

> plus my

hot water heater sidearm BTU transfer rates?

Probably not, though there are situations where you do.  Are there any large whirlpool baths or multi-rainhead showers?

The zone valve should be specified with a Cv value.  This will allow you to calculate its head loss for a given flow.
• Posts: 6
RE: first things first

So what you are saying, is that I should use 39,0000 max capability of my ft. of fin coils transfer capability to calculate my GPM.  This would be then  39000 divided by 9000 using my above equation.  So My GPM target would be  4.3 GPM.

If my new target is 4.3 and my pump can pump 4.3 against the 13.3 head loss, how do the individual parallel paths figure up.  If each path needs 4. GPM to get the  2-4 ft/sec and according to earlier posts (unless I am confused) It seems that to deliver 4.3 to each of 4 parallel branch, you would need a pump that could deliver 4 times 4.3.  I am having a hard time with the relationship to GPM capacity of the pump with the parallel loops.    If a pump can deliver 4.3 GPM with this same flow fill 4.3 GPM requirement through 4 parallel paths?  Hurry up "pumping away", "parallel loops" books I have on order....

I am doing a heat loss calc ... or at least a friend of mine is.  I gave him the numbers today of my layout and R-factors, windows, doors, facing directions  etc.

There is a large whirl pool bath, but we hardly ever use it.  In current config, we cannot fill it up 1/3 before we are out of hot water.  I have a amtrol 7L 41 gal tank.

Zone valve has an equivalent length of 2.5 ft  of 3/4 copper so it is negligible in my current calcs.

I understand that the 007 can't even overcome my head loss so I figure that is where the boiler temp rise is coming from, but  would someone please give me a more detailed relationship of the pump GPM versus the GPMs needed in parallel branches.  This is still kind of foggy.

Thanks,

Taggs

• Member Posts: 139
edited August 2012
GPM through parallel branch circuits

I think that if your circ pumps 4 GPM through one zone and another zone valve opens up, your circulator would see 8 GPM but each zone would get 4 GPM (after they branched off) if they were the same size.

This would mean that if circulator could pump 4 GPM into the largest zone when it is the only one open, that the circulator would be large enough to get the same flow to the smaller zones as well unless there is more restriction in the smaller zones.
• Member Posts: 139
Indirect

"There is a large whirl pool bath, but we hardly ever use it.  In current

config, we cannot fill it up 1/3 before we are out of hot water.  I

have a amtrol 7L 41 gal tank."

I have not looked into this tank/indirect or whatever it is.  You may need more flow to get a better recovery rate.  If you look into this, I am thinking you will find that you need a 0010 instead of a 007 to get the optimum recovery rate.
• Member Posts: 9,514

Head is calculated using the highest head loop. Flow requirements are calculated using the flow rates needed in the loops. So you need a circulator that will give you the required gpm , and be able to overcome the highest head loop. Example if your head is 11' in your longest loop, and you have 4 loops requiring 4 gpm each then you need a circ that will deliver 16 gpm at 11 ' of head. If you are zoning with zone valves you will need a differential bypass so a single zone does not try to get full flow.
• Member Posts: 299
Heatloss

Remember the flow in the zone must only match the heat loss not the connected load. If you have 39,000 btu's of radiation but the heat loss is 20k than the flow rate needs to match the needed heat requirement.

If you size the pump for the connected load than the system flow rate will exceed the flow rate required in the boiler. The flow rate in a cast iron boiler normally should be between 20f and 40f delta-T. If the boiler matches the heat loss of let's say 60k but the connected load is 90k that means the boiler which should have between 3 and 6 gpm will now have 9 gpm flow. Delta-T will be less than 20f.
• Member Posts: 2,144
edited August 2012
Heat Loss

I'm confused. The  btu load is 39k ?  The total btu emiters is also 39k? So KindaSorta 19k(2GPM),10k(1GPM), and10k(1GPM) zones? That's only 4 GPM or so for the total house ,excluding the sidearm .?????? And when you tested  there were two zones and the DHW on line at the same time?
• Posts: 6
This makes logical sense to me

Logically, it seems that I would need to overcome the worst head loss and if all zones were open and I needed to keep my flow rate between 2-4 ft/sec (4gpm) in 3/4 copper pipe, the circ pump would need to be able to supply 16gpm if all zones were open.

I have 4 loops all 3/4 copper:

Loop 1 120 L and baseboard that will give off 8,800 btu when figuring 550btu per foot @ 180 degrees

Loop 2 234 L and baseboard that will give off 11,500 btu when figuring 550 btu per foot @ 180 degrees

Loop 3 277 L and baseboard that will give off 18,500 btu when figuring 550 btu per foot @ 180 degrees

Loop 4 is the hot water heater and has 106 L

Main Loop supply is 1.25 and is 117 L

4 zone valves come off the Main loop supply feeding the zones

Lochinvar has built in bypass pump that is on whenever the circ pump is on.

Contractor installed a Lochinvar Solution 110,700BTU output and is using

Circ pump is Taco 007

Pump is pumping towards the pressure tank which is about 20 inches down stream from the pump. water feed and pressure valve is located 20 ft from the pressure tank and is attached directly into the 1.25" main loop supply.

Nothing works well. I am trying to straighten this mess out.

I do not know my BTU load, yet. It is being calculated. For now, I would like to make this system as efficient as possible as is. I will have to wait until I can afford to add more baseboard should my BTU load come back way more than what I have feet of radiators to provide. I paid a boat load to the contractor to install what is here. This is the reason that I decided to learn this stuff on my own so I will know better next time.

I like the idea of the differential bypass. I am thinking about a programable one that will key off of the opening of each zone valve. Would this work?

Thanks,

Taggs
• Member Posts: 3,515
Bypass

You wouldn't need a bypass if you used either a Grundfos Alpha pump or the Taco Bumble Bee. The variable speed pump would provide the correct flow for any zone and reduce speed as the various zone valves close. I'd also move the pump to beyond the expansion tank, so it could be pumping "away". That will eliminate the 8# pressure differential.
• Member Posts: 4,469
This discussion has been closed.