Circulator question

Mark Wright

I am trying to educate some young folks on circulators and their pressure difference.   We have piped an expansion tank, followed by a Taco 007 and a flowcheck.  Out of the circulator is 3/4' pipe to a bullhead tee, going to 2 exact length pieces of fin tube.  This circuit also has a baseboard tee with a coin vent.  The supply water is through a Watts 1156 boiler feed valve piped to the expansion tank and set at 12psi.  We bled all the air out......and when we started the circulator....expecting an 8# difference across the circ we got much less....like only a 1 to 2 psi difference.  What did I do wrong????? 

Mark Wright

pumping away

we piped it pumping away from the tank of course

Mark Wright

pumping away

we piped it pumping away from the tank of course

Charlie from wmass

I do not see how you expected

such a large pressure differential? You pumped away from the expansion tank also known as the Point of no pressure change.

Mark Wright

I have always thought???

That you would get an 8 psi Differential on  a circ in this situation.  If you read Dan Holahan's book on pumping away he insinuates that you would see an 8psi differential in this situation.  I am just a simple hvac inbstructor with most of my background in ref/ac.  So was i expecting too much from this project????

Charlie from wmass

open the flow check and reverse the circulator

It should be a good example of why you should not pump towards the tank. How long is the total circuit?

Mark Wright

i did have them open the flow check but still pumped away

We did not get any better results.....why?

Mark Wright

It ios just a piping project...no boiler

The curcuit is only 2 pieces of baseboard and maybe a total length in piping of 10 ft

Charlie from wmass

You did not get

a large pressure differential because you pumped away from the expansion tank. That is the reason for pumping away. Take the pump unbolt it and flip it so it pumps towards the tank. Then check back tomorrow night with the results.

Charlie from wmass

Mark do you have photos

of the set up you had? Also remember there is an assumption of much more head on a system than what you have.

Mark Eatherton

In order for there to BE a difference, you must MAKE a difference...

Sounds like a short loop situation. One thing you can show to the students is that based on the noted differential in pressure, converted to feet of head, the circulator under THAT condition is moving "XX" GPM. Throw a ball valve down stream of the pump and choke the flow, and you will see a greater pressure differential.



Take THAT pressure differential, and again, convert it to feet of head, and using the manufacturers performance chart, show them how many GPM's the pump is moving under THAT condition.



Also, I realize that this is just for demonstration purposes, but I think you should explain to your students that they should never use bull headed tees in their practice in the field.

The two exceptions to this rule, are primary secondary piping, and pump prioritized applications.



The conversion from PSI to feet of head is to take the PSI, and divided by 2.307. In other words, 1 PSI = 2.307 feet of head.



Different systems have different pressure drop for a given flow. Make the student do the math so that they will then own it.



Let us know if you have additional questions. We're here to help and to educate.



ME

Mark Eatherton

Sorry Charlie...

But reversing the pump is only going to show a negative differential for that given scenario. You are correct that he is not creating enough friction in the piping to see much differential.



This is the reason the manufacturers generate pump performance curves.



Low PSI differential = low head, = high GPM flow.



High PSI differential = high head, = low GPM flow.



Think about it. How many GPMS will a double ot 7 move in a large bore piping system like an old gravity system compared to an application where it is seeing small bore, say 3/8" pipe?



Pumps are basically pressure differential machines. The amount of differential depends on what the system does with the flow and pressure drop.



ME

bill_105

The bullhead thing

ME, could you rant about the horror of bullhead tees' I just don't understand why not to do it. I haven't frozen to death yet.I have a few more boilers out there with the same. I expect it's an engineering thing. Just dunno?

Mark Eatherton

Caution..... Educational Rant ahead....

Bill, One of John Siegenthalers very first articles had to do with the use of bull head tees in a closed loop circuit.



It causes hydraulic imbalance because the direction of flow, and subsequent pressure drop, which affects flow continually changes. Not good.



Lets say you have 10 GPM going into a bull head tee, with each branch of the tee requiring 5 GPM. As the water starts flowing initial, it seeks the path of least resistance. As more water starts flowing through branch A, the pressure drop in branch B becomes less, and it becomes MORE in branch A, so the flow slows in branch A and increases in branch B, until it's pressure drop increases, then it starts all over again.



Remember, water is wet lazy and stupid, and has a really hard time making up its liquid mind. If you had balance cocks on each branch, you MIGHT be able to somewhat stabilize the flow, but it would be tricky. The code use to disallow the use of bull head tees in hydronic systems, but due to inspector confusion over primary secondary piping, and pumped priority, they abandoned that clause of the code completely 2 code changes ago.



The ideal way is to split flow to two branches by drawing from the run and the bull. This way, the hydraulics "lock in" and flow remains stable, until interrupted by a zone control device. I've actually felt the back and forth surging associated with bull head tees using the OHHhhh AHHHhhh method. If you had flow meters in each branch you'd be able to physically see it.



WIll bull heading work in the field? Obviously, you know it does. Remember, there is a LOT of room for slop in hydronics and hydraulics.



Grpahically speaking, it looks good and makes sense. In reality, it's a bad piping practice (with the noted exceptions).



Some water heater manufacturers (numerous really) use a pyramidic manifold design for multiple tanks, instead of parallel reverse return piping. Old habits are tough to break. Truth be known, a pyramid parallel uses more soldered joints than a properly piped parallel reverse return...



Hope you enjoyed the rant :-)



ME

Mark Wright

Thanks for all of your responses!

I do believe that my problem is too short of a run and the ball valve idea, creating resistance should have come to me, but didn't.   I only bullheaded the system to keep it compact and manageable as a moveable teaching platform.  So I will try the ball valve Idea and see how it works.

Many thanks to those who responded.  I do appreciate your time and expertise.  This wall is a testament as to why tradespeople are the best!!

LarryC

Additional suggestions.

This is coming from a sparky who likes hydraulics, so take them for what they are worth.



In addition to adding a ball valve on the discharge of the pump, to increase system head, consider adding a ball valve on the suction  side of the pump.  Then you can demonstrate the difference in centrifigul pump behavior when throttling on the suction versus throttling on the discharge.



I assume you have multiple pressure gauge ports in order to demonstrate the pressure changes between various locations of the system.  Can you add an additional expansion tank with isolation valves to demonstrate the difference between pumping to and pumping away from the point of no pressure change?



By pumping to the PONP with a significant restriction after the expansion tank, can you drive pressure at the suction of the pump to below atmospheric?



By valving off both expansion tanks, you can demonstrate the power of a "solid" system where small changes in temperature can create large changes in pressure.  Obviously the relief valve must be in working order.

Unknown

Delta P Across a Circulator

Guys, I must be missing something here.



The pressure differential across a circ (pump), or delta P, or boost pressure (whatever you want to call it) on a closed loop system is based on the pump's performance curve and it is the pressure where the system curve intersects the pump's performance curve.



And, this "boost pressure" needs only to be high enough to overcome the circuit friction loss at the flow.  A 007 providing 2 PSI differential pressure (x 2.31 to convert to feet of head) will provide about 18 USGPM flow (enough flow for a 180,000 BTU load at 20 deg F temp differential.



Even if the Taco 007 was dead headed (run at zero USGPM) the differential head (or shut off head) is 10.5 feet or 4.5 PSI.



I don't know where the expected 8 PSI differential comes from.  Remember, the expansion tank is only there to compensate for the expansion and contraction of the system fluid as it's temperature changes (that's why it's called a thermal expansion tank) and 100% for sure has to be on the inlet side of the pump.

Mark Wright

My 8 psi differential

Thanks for the info, i am going on the reading in Dan Holahans pumping away book.  I'm not an engineer by any stretch just a HVAC teacher in a technical high school.  So no laughing too much.  In my recollection of the reading it gave an example that the differential changes according to resistance and the relationship to the point of no temp change.  For instance if you are pumping away form the pressure tank and assuming the make up water is at the tank(12 psi).  This would then transfer the diferrential to the discharge side of the circulator thus resulting in a 20 psi discharge (12psi inbound supported by the boiler feed plus the "8 psi" differential from the reading). 

And if it was pipped on a return then whatever the inbound pressure "say 3psi" then the discharge would be 11psi.  And if you are pumping at the point of no pressure change where it "cannot change" then the inbound would decrease by the theoretical "8psi".  So 12psi minus 8 will result in a 4psi inbound.  If, of course I read and understood it correctly!  School me further if I am wrong.  Thanks  

Unknown

Pump Training Materials

Mark, send me an E mail with your mailing address.  I will be happy to send out our training information (systems, pumps - the whole deal).



And feel free to use in your classes!

Mark Eatherton

I don't think you're missing anything...

You may have some loose screws, but it looks like they're all there ;-) (Said with tongue FIRMLY planted in cheek)



I just don't have every manufacturers performance curve indelibly etched into my memory like you do :-)



I think we said the same thing, just differently. Different roads to the same point on the map.



Thanks for chiming in, ol' Canadian/Deutch Pumpen Meister!



ME

Mark Wright

Many, many thanks

Thanks to all who have offered their knowledge and wisdom to help me reach my answer.  It is certainly nice to know that so many will offer advice here and ask for nothing in return.



regards to all..........mark wright

jonny88

ME.

when are you publishing your book.i attenended a siegenthaler seminar and was lost after 2 hours.i like dans books for there simplicity and always enjoy reading you comments