water hammer in low return

D2

Hi all,

   I work for an HVAC contractor in lower NY State. Upstate for those of you in NYC. We don't see a whole lot of steam, mainly single pipe in older homes and I thought I had a pretty good grasp of things until we were asked to look at a building one of our customers owns with a fairly large steam boiler feeding several apartments.



   I guess I'll start with the service history since we were called in to look at it. Then I'll give as much info as I can on the piping arrangement. This is a bit long as I'm trying to include as much info as I can since I've obviously overlooked something. Also, it's in the order that I got the info myself.



  We were originally called to install a properly sized relief valve and also a manual reset pressuretrol. The pressuretrol wiring is all line voltage and there is no switch near the boiler and the electrical panels (all 5) are not labeled so I opened a junction box and unwired the boiler to install the pressurtrol. At this point, I didn't really look around much, just installed the relief valve and the pressuretrol, rewired the boiler and beat feet.



  The next night we were called for no heat and the building owner assumed I had mis-wired something. As it turned out, everything was wired correctly, there is a McDonnell #61 float type LWCO being used to operate a boiler feed pump. This was full of mud, and after it was cleaned out everything worked properly. Or, the way it did before anyway.



  I haven't been back for about  6 weeks until a week and a half ago when they had no heat again. This time, the building owner said boiler was firing, but they had no steam in the radiators. He went around checking all the radiators and when he made it to the second floor he removed one of the steam vents and water came shooting out.  He says they drained the boiler and waited for it to refill and they had heat after that and he called us to find out what went wrong. When I got there, they had heat, a normal waterline, and steam shooting out of the overflow on the boiler feed reciever as well as some water hammer in the return line. I told the building owner that I believed the F&T trap wasn't working and I'd look into it and get back to him.



   There is a Dunham-Bush 30-7A F&T trap just before the inlet to the boiler feed reciever. I checked with State Supply, ordered a repair kit, informed the building owner, and three days later installed the kit which was just a new float, thermostatic trap element and a new cap for the thermostatic trap since this is a Barnes and Jones repair kit. As I was installing the repair kit, I noticed the float was indeed siezed open, and also that there was a coupling leaking on the return, which is lower that the F&T trap. Someone used the shipping couplings to assemble the return and this one had rotted out. The threads were pretty well gone on the ends of the return piping originally connected by the shipping coupling so I left that, and went back the next day and  put a 2" Dresher coupling on it. I boogered that up and that leaked too. A real steady drip that was almost a trickle. I  left it for the night and found another 2" Dresher coupling the next morning, went by, put it on successfully, ran the boiler for awhile until the return got nice and hot and I could tell it wasn't leaking, and then noticed the water hammer starting in the return and the steam coming out of the vent on the boiler feed receiver. When I saw that, I thought leaving the leak might have ruined the thermostatic trap with the water hammer in the return, but the return makes a vertical 90,runs about 18" then a horizontal 90 before it runs into the F&T trap so I'm not so sure.



   This is when I started really looking around the basement. The boiler was installed two years ago. It's a Weil-Mclain 1,840,000 BTU boiler. Could be a 1,084,000 BTU boiler though, I'm doing this from memory since we have ice outside today and that info is safe at the office. I can get it if it becomes important. I've never been up into the apartments so I don't know what's going on up there except that in the basement it's piped as a single pipe system with the mains pitched away from the boiler to a wet return.

   

    There is a single pipe coming out of the side of the boiler, 4-6",  makes a 90* turn towards the ceiling for about 24", then another 90 across the top of the boiler, one more 90 towards the floor and then it reduces to 2", and about a foot below the water line, a 2-3/4" reducer where the boiler feed line is connected. The "Hartford Loop" is taken off of the 2" section of the header drip, runs down to the bottom of the boiler, makes a 90 and runs about 3 feet and is connected to the front side of the bottom of the boiler.



  There are tow T's in the header and two 4" mains running opposite directions. On of the mains has another T where a two inch run comes off and heads towards the back of the building.



  The mains are insulated about 4' away from the boiler, the insulation being asbestos and they had it removed from the near boiler piping when the boiler was installed two years ago. Most of the take-offs come off the main at a 45* but there are a few that come right off the top with a 90. I measured off of the ceiling beams since the floor is pretty uneven, and the bottom of the lowest steam pipe is 15" from the ceiling. The water line in the boiler which is about 60' away is 47" from the ceiling so it looks like 32" of "A" dimension to me.



   There are two mains, one runs about 60 feet, one about 80 feet. Only two 90's on each. There are main vents at the ends of the mains, both piped the same off of a  T just before the 90 where they run down to meet the return. One of the mains is piped like this at the end : T with a main vent, close nipple, 90* down, 18" or so vertical run towards the floor, 6 feet horizontal run, 90*, 24" or so vertical run into a T into the return. On one side of the T is the return from the second riser, the other side runs about 8-10 feet then through a poured concrete wall. The insulation continues from the main all the way along the return and stops at the wall. Where the return from second main enters the T where the first main ties into the return, there is no insulation all the way back to the second main.



   The second main has a vent at a T just before the 90* down into the return.This one runs straight down into the return, which runs about maybe 10 feet at most into the T where it meets up with the insulated part of the return off the first main. The return runs about knee high about 60 feet across the basement, 90* up about 18", 90* horizontally about 4 feet into the F&T trap, out of the F&T trap about a foot, 90* down about a foot, 90* horizontally about a foot into the boiler feed receiver. The feed line from the pump is 3/4" copper running across the floor 12-15 feet,90* up about 18" into a 2" - 3/4" reducer into the header drip.



    I can provide drawings with dimensions if need be, Actual boiler size, and near boiler and feed pump piping details if needed. If I put too much extraneous info in here I apoligize, it's definitely a long read.



   The first thing smacking me in the face here is whey there is steam in the return at all. It should be full of water. I have looked around the basement and don't see any signs that it's leaking except where I put the Dresher coupling on it. The second thing I'm wondering is why there is an F&T trap at the inlet to the boiler feed receiver since the returns should be full of water and I don't see any signs of leaking. The water hammer in the return isn't real loud. Sounds like someone tapping on the return with a small tack hammer every now and then, say, once a minute or so? A few minutes after that steam starts coming out the receiver vent, that continues for a few minutes, stops, all is quiet for a few minutes, then the cycle starts again. I would assume that after 24 hours, the return should be full of water, there should be a puddle on the floor, or there should be some real loud water hammer in the mains. If it takes a long time, I may have missed the water hammer in the mains but noone has complained about it. Then again, they may be used to it as a fact of life by now and I don't get to talk to them so I suppose it's a possiblity.



    So, if you've gotten this far any insight will be gratefully accepted, and if any furthur info is needed I'll get it as soon as I can. Thanks to any and all input, opinions, etc.



  Dave



  



  

BobC

Pressure?

I'm just a homeowner with a single pipe steam system but I was wondering what pressure is the boiler running at? If it's too high it will find it's way into all kinds of places it shouldn't. Be careful, that 30PSI gauge might be lying to you.



Are those vents large enough for that size and length of pipe/



Bob

Jamie Hall

Pressure

It is well worth keeping in mind that with an F&T trap and a receiver, your returns are at atmospheric pressure -- always (unlke a straight piped gravity return, where the returns are at boiler pressure -- always).  In the situation which you have, excessive pressure in the system can push steam down the drips and into what should be wet returns.  At first glance, it would be my estimate that this is what is happening to you.  Ideally that 18" up into the F&T might be enough to provide a water seal in the wet returns -- but 18" will only balance perhaps 12 ounces per square inch, and that possibly not reliably.  So it is quite possible that steam is pushing past that water seal...



That's just a first thought...

D2

vaporstat?

The building owner called again, more squirting radiators. I went back and he wasn't around and there was a lot of water in the basement where they drained the system.

I watched the boiler for one full cycle which lasted half an hour. There was no steam coming out the vent on the boiler feed reciever, but once the boiler got steaming which took about two minutes, the water line bounces so bad the pump feeds about once every minute or two for about 3 seconds for the rest of the call for heat. I'm wondering if the contacts on the float switch aren't hanging up.



   It took a half hour for the condensate to get back to the receiver. I went around to all the apartments and the two stores and checked the radiators. The stores are on the first floor with two apartments on the second and two more on the third floor.



     There were three out of four radiators in an empty store on the first floor where the carpet was soaked but these stayed dry during the firing cycle when I was there although there was a little banging from one of them. They were pitched slightly away from the radiator valve, and on two of those the radiator valves were just about closed.

   

    In the second store, all dry, all hot, no banging, vents worked and most of the valves were just about closed.



   In one of the apartments on the second floor  they said the radiator vent spit and this was on the same end of the building as the wet radiators in the store. Most of the radiator valves were just about closed, most  radiators were pitched away from the valves slightly, about half the vents didn't open or didn't close, about half of the packings around the valves were leaking. Tightened the packings, opened the valves. Same in the two apartments on the third floor but no complaints of vents spitting.



  In the basement, from the bottom of the lowest main to the return is only 28", so with the 18" rise up to the F&T trap at the receiver that's only 10" to get condensate back into the receiver right?



    The pressuretrol is set for .5 psi on and 1.5 off. The boiler is actually 1,040,000 btu's in, 842,000 output, 2633 sq.ft. of steam. There is only one riser coming off the side of the boiler, 4". The header is 4", and the main is 4".  The Lost Art book says 5" header for that size boiler.  The main runs 10' then T's into 2 mains, both 4" for about 40' on both, then reduces to 3" with an eccentric reducer. One main is 80' total one is 70' total length. 4 90's on each, one 90 and the T on the main before it splits. When figuring friction loss, do  you count all the T's for the radiator runouts too when figuring TEL right? If so, one main ends up at 340', one at 260'.



    I'm wondering if a vaporstat might help with the water hammer in the return? And also, in every piping diagram I see, the F&T trap is at the end of the mains, not right before the receiver for the boiler feed pump. What is that doing there?



    Thanks for reading all this. I hope I've provided enough of the right info to get a better look at this system. Thanks again!

 Dave

   



  

jpf321

please look hard ...

and truly justify and understand why there is a feed pump and F&T trap at all, what is it's intended purpose and what is the consequence of having it.



i just recently assisted in a school (see thread: http://is.gd/4SUK7z) where they suffered from severe hammer in the returns because there was a pump and F&T's which were keeping the floor returns DRY .. we cut out and abandoned the F&T's and the pump and everything got better immediately with a normal gravity return to a false-waterline into the boiler.



also,

+ is the hammering on the boiler side of the pump or the far side of the pump? 

+ does the hammer persist through the entire cycle?

+ is there any part of the cycle where it is better/worse than other parts?

+ does the hammering coincide with the pump operation at all? when pump is full? empty? pumping?

+ is the pump properly vented to the air?

+ when you say it came from pump overflow what does this mean?

+ is this a true feed (on boiler request pump)? or a condensate return (when full dump) pump?



of course pictures would be really nice of everything as well.



where "Upstate" is this? perhaps I could come for a look.

D2

Pic's

Here are some pic's of the job. I don't know how they're going to show up or if you can enlarge them but at the moment they are to the top of this post and pretty small.



The first pic is the boiler feed receiver, you can just see the top of the motor in the lower right, and the return is the 2" pipe just below the waste line. The pipe you see in the front coming off the receiver is the vent/overflow. This is wired to the McDonnell type 61 LWCO in the third pic there. The 61 is located a few inches below the waterline at startup. When the waterline drops the float drops and kicks on the feed pump. On startup, a few squirts of water come out the pipe T'd off the reciever vent pointing down in the first pic. A half a quart to to a quart maybe. This happens 2 or 3 times at about 1 minute intervals for the first 2 or 3 minutes after the boiler fires then stops. The float in the F&T trap was seized up and there was steam pouring out the vent before we rebuilt it.



   I'm not sure why that receiver and F&T trap are there. The lowest steam carrying pipe is 15" from the ceiling, the waterline is 48" from the ceiling so we have 33" of dimension A assuming level floors. They appear pretty level with my head up in the floor joists looking across the bottom of the joists but it's an old building and a water level would be in order if we were going to consider removing the feed pump.



   I watched the steam cycle for an hour when I was there on Wednesday. The return was cold, the steam cycle lasted for half an hour and the return didn't start to get warm near the receiver until the end of the cycle. I could hear condensate gurgling in the F&T trap 25 minutes into the 30 minute cycle. The return is 70' long, the basement is usually 65-70* .The feed kicked on 13 minutes into the steam cycle due to the bouncing waterline, at no time during the steam cycle did the waterline get below the sight glass. My assumption is that the receiver was put there due to slow returning condensate at some point but I'm here asking questions because I really have no idea. I'll list my assumptions just so wrong thinking can be corrected. I can tell you that when had the rotten coupling off the condensate was running out of the return pretty freely, pretty hot, and emptied the return in about 5 minutes. It was also very clean.



    In the third pic is the end of main l, and you can see the return is insulated. In the fourth pic is the insulated return from main l all the way down to the common return. The common return runs to the right. Just on the other side of the T there in the common return, where it runs through the wall, is the end of main ll in the sixth pic.

Where the insulation on the return from main l meets up with the common return, the insulation continues for about 14' on the common return, although it is not insulated from the T back to the end of Main ll. I'm wondering how much the insulation is contributing to steam getting into the return line.



    What has been happening is that the building owner calls us after he has water squirting out the radiators, drains the system from the drain bib on the receiver and the boiler, refills the boiler and the receiver, then calls us. He's assuming the pump keeps feeding the system too much. I've never witnessed this, have taken the lwco that the pump is wired to apart and it's clean, it works, doesn't hang up. So I'm not sure if that is indeed the problem or if he has some other issue.

  

   + is the hammering on the boiler side of the pump or the far side of the pump? 

   + does the hammer persist through the entire cycle?

  





+ is there any part of the cycle where it is better/worse than other parts?

 

The water hammer starts at the end of the steam cycle, a few minutes before the boiler shuts down, and for a few minutes after.





  It's in the return that runs across the basement about knee high.  This, only after the system has been drained from what I've witnessed  so I'm assuming the returns are empty and it makes sense that steam would be getting in there with no water since the F&T trap is at the receiver.



+ does the hammering coincide with the pump operation at all? when pump is full? empty? pumping? 

   I would assume it's full all the time. There is a float in the receiver that keeps it full, I have checked it and it is working.







+ is the pump properly vented to the air?

    In the first pic of the reciever, the pipe that is 90'd down to the floor is t'd into the vent pipe. It's about 5 feet high and they have a length of plastic corrugated pipe connected to it running to a sump. They say water has come out of it but I've only seen steam coming out of it.







+ when you say it came from pump overflow what does this mean?

   That would be the pipe 90'd down towards the floor that it T'd into the vent on the receiver.







+ is this a true feed (on boiler request pump)? or a condensate return (when full dump) pump?

   It's wired to the McDonnell 61 float type LWCO that you can see in the third pic there. To the left is the LWCO that shuts the burner down which is located below the sight glass.



   Thank you for your time, I hope I haven't confused the issue too much. I'm pretty confused myself. The job is in Middletown, NY. Probably an hour at least from Queens. We only get down to the Bronx, Mt.Vernon and Yonkers on a semi regular basis so I'm not exactly sure how far Queens is but the job is about 45 min. form the Tappan Zee bridge with no traffic.  If your "consulting" fees are withing my budget, I'd love someone who knows what they're looking at to check this out. "-)

jpf321

pics?

I don't see any pics.

nevermind......you got them up now

jpf321

a few top of head thoughts...

+ the return and equalizer may be bushed down to small according to boiler spec

+ the 1" copper from pump to boiler may be too small

+ is there an autofeed feeding the system with freshwater somewhere?

+ can you step back and take another pic or two of the pump

+ if the pump is on request of the boiler, then it's entirely possible that the boiler is not requesting water and it is overfilling the pump right out the vent pipe (that is a VENT pipe and is not an OVERFLOW pipe)

+ the boiler header doesn't look to be proper height above waterline as per boiler spec

+ unless there is determined to be a significant need, or it is required according to boiler I&O manual, it may be wise to consider elimination of the pump.

+ the 2" returns may be too small for the system

D2

Top o my head responses

+ the return and equalizer may be bushed down to small according to boiler spec

        They may very well be. The main heats up all the way to the vents pretty quick when the boiler starts up though, it that has anything to do with anything



+ the 1" copper from pump to boiler may be too small



+ is there an autofeed feeding the system with freshwater somewhere?

  Only to keep the receiver full



+ can you step back and take another pic or two of the pump

   If I end  up going back I could. You can see the F&T trap there, and the return line behind the large waste line. The copper pipe coming off of the tank vertically is for the feedwater to the reciever. You can just see the top of the pump motor in the front and that's all there is to it.



+ if the pump is on request of the boiler, then it's entirely possible

that the boiler is not requesting water and it is overfilling the pump

right out the vent pipe (that is a VENT pipe and is not an OVERFLOW

pipe)

The pump comes on when the float  in the McDonnell 61 LWCO drops. This happens almost every minute or two after about five minutes of firing. The water line bounces like crazy and is giving the feed pump a real workout.



+ the boiler header doesn't look to be proper height above waterline as per boiler spec

 Probably isn't. The riser is 90'd off the side and there's 23" center to center on a four inch pipe from where it comes out of the boiler to header. The waterline is probably 8-10" below that.



+ unless there is determined to be a significant need, or it is

required according to boiler I&O manual, it may be wise to consider

elimination of the pump.

   The lowest main is 15" from the bottom of the floor joists. The boiler water line is 48" from floor joists. So that would be 33" A dimension right? You might lose 5 or 6" across the basement due to the floors not being level anymore so it seems to me we're real close to not needing the pump at all unless the returns are real slow.

   The return does run over 70' back to the boiler. I don't believe you have to pitch that though, do you? If not, I don't see the pump is even there. But then, there is a reason I'm asking questions of you folks."-)



+ the 2" returns may be too small for the system

   I have a chart here that says 2" returns are good for 4,000 sq.feet EDR and 960,000 BTU's. The boiler is rated at 2,633 sq. ft and 840,000 btu output. However, someone may have checked the wrong box on the boiler tag.



    I do appreciate the responses, and hopefully the problems will continue so we'll be forced to get to the bottom of it. But noone was complaining last year and the winter is almost over so I'm sure we'll be doing battle with this one come next winter.

We just went out and changed the float in the LWCO/boiler feed  so if the problems continue we'll have to start looking at the piping. There's nothing else left to replace.

Steamhead

That's

a Weil-McLain LGB series boiler. It looks like one of the smaller ones, so the 4-inch piping should be OK. Double-check to see what the actual model number is (LGB-something) to confirm this.



Not sure why that return pump is even there. But the fact that there's only one trap, which is in the wrong place, tells me why you're getting water hammer. That return was originally filled with water, and the water returned by gravity. The steam could not enter the return since the water was in it.



But with that pump and tank, the water does not stand in the return as it did. So the steam enters the return, where it bangs.



Get rid of the pump and tank, and pipe the returns to a Hartford Loop and it should work fine.



LGB boilers come standard with 2-stage gas valves. You can hook up a Vaporstat to drop the burners to low fire after the steam pressure reaches 4 ounces or so. When set up properly, this will result in a long, steady burn and save some gas.



The main vent in the pic looks like an ancient Dole unit. It's probably way too small for the main (if it's working at all). Measure the length and diameter of each steam main and we can tell you what you need.

jpf321

aha .. now maybe we are getting somewhere ...

the autofeed shouldn't have to "keep the pump full" .. the pump should fill from the return water.



when the boiler calls for water, does it completely empty the pump?

when the boiler calls for water, does it immediately trigger the autofeed?

is there a delay at least on the autofill?



it seems like, perhaps:

1) due to low header height, the water is jumping out of the boiler and perhaps wreaking havoc with your waterline

2) the unstable waterline is triggering the 61 to call for water from the pump

3) the pump is sending water when in fact, with a stable waterline, the boiler wouldn't need it

4) the autofeed on the pump is filling the pump prior to the return water getting back to it (again perhaps due to unstable waterline, and the waterline is unstable due to low header height)

5) you have now introduced extra water into the system by the time the return condensate gets to the pump, the system begins to overfill.



on the next cycle, presumably, the pump is certainly full. when the 61 calls for water, depending on how the autofeed is setup, you may introduce even more water.



couple the above presumed cycle(s) with perhaps an undersized F&T, or constricted returns, and you may indeed get overfilling eventually.



i would see if the autofeed has a delay, and/or turn the autofeed off for a cycle or two. presuming that your LWCO(s) are functioning properly, you shouldn't hurt anything. You do have manual reset on your LWCO(s) somewhere right? that could pose a problem during this test.



+ check all the installed heights of the 61 and the LWCO(s).

+ check the autofeed float installed height

+ determine if the boiler needs any skimming.

+ we already know that the header (which should be at least 24" above NWL, ideally 24" above jacket) is too low, consider a drop header repipe.

+ determine if the autofeed has a delay and enable it

+ consider abandoning the pump

+ consider a larger capacity F&T

+ consider installing a feedwater meter (Istec 1700 or 1710 http://www.istec-corp.com/pdf/1700EngManPLR.pdf)

D2

Main vents

There is 6' of 4" pipe off the header to a T.



Off of this T: Main #1 runs 31'  @  4"

                           reduces to 3"

                                     runs 44.5'@ 3"

                  For a total of 75.5' 



The mains are pitched down from the header



                       Main #2 runs 34.5' @ 4"

                                reduces to 3"

                                      runs 30'  @   3"

                        For a total of 64'



Just to furthur complicate matters:

 off of Main #1, 50' off the T, there is a 1 1/2" run for 34' pitched UP from the main,



 reduces to 1" and is pitched down for 20', vents, then drops 57" and ties into the return just behind Main #1( see pic.)  Total length of 54' off Main #1, 104' from T



Edit:  The mains are insulated with asbestos except right around the boiler where it was removed when the new boiler was installed about 2 years ago. The 1 1/2"  run off of  Main#1 is un-insulated .

D2

Main vents

The boiler is a WM LGB 9 

There is 6' of 4" pipe off the header to a T.

 

Off of this T: Main #1 runs 31'  @  4"

                           reduces to 3"

                                     runs 44.5'@ 3"

                  For a total of 75.5' 



The mains are pitched down from the header



                       Main #2 runs 34.5' @ 4"

                                reduces to 3"

                                      runs 30'  @   3"

                        For a total of 64'



Just to furthur complicate matters:

 off of Main #1, 50' off the T, there is a 1 1/2" run for 34' pitched UP from the main,



 reduces to 1" and is pitched down for 20', vents, then drops 57" and ties into the return just behind Main #1( see pic.)  Total length of 54' off Main #1, 104' from T

D2

found the boiler manual

For whatever reason, I can't seem to respond to this post. I keep gettting redirected to log in after I've already logged in.

 I found the boiler manual. They spec a 5" header and riser and 2 1/2" equalizer. Would that make the header height too short? They only want 24" above the water line and there is more than that.

  The 61 Is the boiler level control/ primary LWCO. There are no delays on that,  or the receiver feed which I believe is just a float.



  

D2

short posts

I seem to be able to respond if I keep it short. This is pretty frustrating.

However, the 61 is where WM wants it. That works everytime I've been there and tested it meaning it's shutting off when the float rises. I'm thinking the water out the radiator vents must be condensate backing up. I've talked to some of the tenants and there's only one radiator on the second floor spitting. it's on the same end of the building where the radiators are soaking the carpets and there is some funky piping under that. Some 1 1/2 uninsulated piping that pitches up from the main which is pitched down away from the boiler. There are only two runouts off of this piping with two more capped.

   All the problem radiators are on this end of the building where the mains end. I'd like to get into the apartments and size all the radiators and see what runouts serve one radiator and which are risers serving several radiators. There are about 5 or 6 capped T's off the mains and from what I've seen the connected load is not all that large.

  I think If we downfire or keep the boiler on low fire or something we could keep the condensate where it belongs and then we could probably abandon the pump when the tank rots out which looks like it could be any day now.

Steamhead

Vent each of the 4"/3" mains

with four Gorton #2 vents on each. This will clear the air out in a hurry!



The short branch main can be vented with a Gorton #1.

Steamhead

The header is too small

and will need to be repiped. 4" is too small for that boiler and is probably making wet steam.



The 24" above the waterline is minimum. More is better.

D2

wet steam

Is there some way I can test for that?