Precharge Expansion tank: old school non-bladder

Keith W.

I hope I'm not being redundant, I couldn't find the answer in a quick search...



I've got a job with a largish residential radiator system retrofit from oil to gas a few years ago. I reused the expansion tank, which is about 30g and is old-school steel tank in the ceiling joists, filled merely with air. It has worked for several years since the retrofit (and decades before), but now the system is overpressured (it's a combi-boiler, so it gets some summer use too), and I have to assess whether the expansion tank merely filled up, or if the auto-fill regulator needs replacement (the only other salvaged part from the oil boiler).



Anyway, I've drained the expansion tank (didn't try to assess if it was full, that was probably a lost diagnostic opportunity, but at near-40-psi there was a lot of water in there), and I could just open it back up to the system now, which I'm sure is what I did the first time (I'm not sure I even drained it before). But it occurs to me that it could make sense to pre-charge it to, say, a few psi below idle pressure (let's say 12 psi), to leave more expansion space... Since these tanks have generally fallen out of favor, I don't see much info on this procedure.



Please don't spend a lot of time trying to convince me to switch to a bladder tank, though pithy insightful suggestions are always welcome. A 30g bladder tank would be expensive and hard to fit into the room. Thanks for any help!

Alan (California Radiant) Forbes

It would help

to know how it's piped into the system, i.e. on the vertical supply coming out of the boiler?



Expansion tanks and their effect on the heating system can be very complex; more than my little brain can comprehend.  I have replaced or recharged many expansion tanks on systems that over-pressurize, only to have the symptoms repeat.  Some have bad fill valves, some have pin holes in the indirect heat exchanger, but the majority seem to just not be able to get along with the expansion tank even though it is properly sized and charged.

SWEI

Not sure I can manage pithy

but (depending on a few variables like system temperature and glycol percentage) a 30 gallon bladder tank (which only holds 11.4 gallons of water in the Amtrol variant) usually belongs on a commercial job with with several hundred thousand BTUs of boiler and several hundred gallons of system capacity.



If you're still curious, try plugging some numbers into the Amtrol commercial sizing calculator http://www.amtrol.com/support/extrol_com_sizing.html then change the model number prefix from an AX to an SX and you have an option.

Steamhead

Also

do you have an air separator which routes system air into the tank so it won't waterlog?

SWEI

An excellent point

first time I ran into one of these I had to trace out the piping and ask around a bit before the light went on.

JUGHNE

Dance with one every 2 years or so;

I have a system I maintain that has the old school tank ( actually it is in an old school); the good thing is there is a sight glass on the tank and makes the level obvious.

If I remember correctly from reading up on old stuff (which is about all I see) is that you need an air separator on the boiler but not any air eliminator on piping or radiation.  The theory being that all the air in the system must go to the exp tank and not leave the piping/system.  Maybe it took this long for a microbubble air eliminator to empty that tank.  BTW the fill valve for this system is off, have LWCO and check nearly weekly in cold weather.

Keith W.

Argh! Lost reply!!!

I just wrote a long reply to your various comments, but because it didn't have a title, it got deleted! Argh! That's a rotten bug in this Wall software, it didn't have to delete my reply! Anyway, here's the gist...



First, thanks for your thoughts and replies.



Second: I still didn't get an answer to the question I tried to ask: has anyone heard of, or practiced, pre-charging a non-bladder expansion tank? Please be specific, so (frankly) I can know how much to weight your response (nothing personal). I have always only refilled the system, letting the developing system pressure charge the tank, which eats into expansion capacity. Another way of saying that is, without precharge, a non-bladder tank is worth a lot less than a bladder tank of the same size, in terms of expansion capacity. With pre-charge they could be fairly similar, I would think (you can never precharge over system idle pressure, of course, with a non-bladder tank, whereas you can with a bladder, which changes the equation).



Now, responses:



All indications are that the tank is roughly the right size, Amtrol calculators concur. I've never heard of an air eliminator feeding an expansion tank, is that really done? And finally, there is no glycol or external heat exchanger, nor reason to believe there're any internal leaks in the nearly-new Triangle Tube combi-boiler.



The tank is plumbed in the ceiling joists of the utility room, with all plumbing entering from below, just as every other one I've ever worked on has been.



Sorry if I'm snippy, I am down one finger for a few weeks following an accident and cranky that I have to retype all this. Thanks again for any help.



Keith

Keith W.

Oops: expansion tank sizing correction

Okay, I try to admit when I'm wrong... I had used the Amtrol sizing calculator before I posted, and then used it again when someone referenced it. However, in my haste I didn't actually check, and made an assumption that an Amtrol ST30, for example, was a 30g tank. I was moving too fast, it's a 14g tank (with 9g of expansion). Which I'll admit is still larger than I've used on most residential systems. Anyway, gone at another way on the Amtrol Calc I got an EX90 (how do they come up with these model numbers??), which is also a 14g tank with 11g of expansion. Both of which might make sense in comparison to a non-pre-charged 30g, which is in fact only an estimate by me looking up at the joists: I didn't measure it.



Hmmm... if I think in terms of Boyles law (p1v1=p2v2), with a very sloppy accounting for the fact that absolute pressure is system pressure plus atmospheric, and rounding like crazy, I'd get empty pressure of 15 psia, idle pressure of 30 psia, and max pressure of 45 psia, which is to say half of the volume would be lost to charging the tank during system fill, and I think another 5 gallons to the next 15 psia... Of course, this doesn't account for temperature changes...

Keith W.

air eliminator feeding expansion tank

Also: perhaps I'm confused, but I don't think an air eliminator could be fed into an expansion tank, I think the back-pressure would keep it from working. Or maybe I'm missing something.

Steamhead

We do it all the time

and it works great- here are a couple examples:

Mark Eatherton

K.I.S.S.....

First off, why 40 PSI? You only need 1/2 PSI for each foot of system elevation above the boiler gage, plus 5 PSI. (Looks like you already knew this)



Secondly, Isolate expansion tank from system and COMPLETELY drain it out, not just to the gurgle, glug glug stage, but COMPLETELY empty. (for the beneit of other readers)



Thirdly, refill system using the 1/2 PSI (OK, for nit pickers, it is .434 PSI per vertical foot) PLUS 5 PSI. Nothing more will do you any good, and in fact WILL take up the acceptance capacity of the tank. Other than the water expansion capacity, Amtrol charts are useless for this type of tank.



In my 35 years of doing this, I have seen ONE system that actually changed the air charge in the tank, and it was the biggest commercial hotel heating / cooling system I've ever seen and they used Nitrogen for the gas charge).



Don't go there. It would be like applying todays automobile ignition technology to a 1950's vehicle. it will not work. Sounds like a great idea on paper, but it will not work right.



Lastly, and you are probably not going to like this suggestion, but it is a proven fact/method of eliminating ongoing air binding/water logging problems. If possible, have all circulators pump AWAY from the expansion tank connection, which I presume is the boiler, and make certain you have an Bell and Gossett AIrtol fitting on the tank to keep the air from getting out. Also, automatic air vents are a NO NO on this type of system.



In rereading your original post (I am a terrible speed reader) You may have another problem, being caused by the new high head pumps and the location of the Pressure Reducing Valve. If the pump is pumping towards the expansion tank connection, and the PRV is connected on the inlet to the pump, every time the pump starts, it creates enough of a negative pressure to cause the PRV to put a little water into the system. Over a period of time, this ends up causing an ver pressure condition. You cold also have a failed heat exchanger bleeding potable water into the system. Turn off the make up, and if pressure stabilizes, you've fond the problem. If pressure continues to rise, you have a failed heat exchanger. The PRV should be connected as close to the expansion tank connection as possible.



Buy Dans book titled Pumping Away and you will learn a lot about the critical relationship of pump and expansion tank locations. Hardest subject I've ever taught...



It's really not rocket science. Rocket science is EASY :-)



ME

JUGHNE

Separate or eliminate air

Thanks, ME, that is what I mean. Pumping Away book page 55 explains this much better than I can.   By my definition an air separator fitting puts air back into the tank.  An air eliminator such as Spirotherm automatically removes air from the entire system and eventually pulls the air out of the plain steel compression tank.

Keith W.

Thanks for replies

Steamhead: I'm not real familiar with steam systems (they aren't used much around here, residentially): that's interesting piping on your systems, some of it is mysterious to me. On the air eliminators I've used, there is sometimes a piping connection for an expansion tank, but it doesn't route the eliminated air into the expansion tank. That's what I was talking about. Of course, since almost all tanks are bladder-type, this makes sense, in addition to being necessary due to the pressure differential problem I think. With bladder tanks, of course, the recommended positioning is precisely to reduce/eliminate air entry. Anyway, I may be misunderstanding your point, and I'm sure I don't completely understand your piping, though it's very clean (are all your basements that clean?) ;-)



Mark, thanks for all that: 40 psi was simply what the system ended up at when it overpressured, though I don't see where I wrote that (I'm an even worse speed reader, and more impatient, apparently). Anyway, you seem to be suggesting to definitely precharge the tank, in this case the height of the system is about 25', so around 15psi should do it. Thanks.



I think I'll just have them operate it with the autofeed off, as a general rule, eliminating the pump-direction/high-head issues, I think. Though I don't think they apply, due to the fact that the expansion tank is sort of far away from the pumps... but I haven't really thought about/looked at that issue. Interesting point.



Thanks for everyone's thoughts. I have never given so much thought to bladderless expansion tanks...