Venting an old vacuum system

Jim_from_Worcester

I'm hoping someone can help me with this.

For background, I am in charge of a 7-story 100,000 SF building in Worcester, MA. Built in 1922, it was apparently a vapor vacuum system from the get go. The original boiler was converted to oil and went off line in 1986 when replaced with a 14 section HB Smith. I have been making incremental improvements for the last three years and the performance has been improving with each season. The condensate pump (which cannot maintain a vacuum), by the looks of it,  was replaced sometine in the 60's or early 70's (It's a Hoffman dual pump unit) and boiler feed tank was also installed (around the same time as the new boiler?).

Problem number one: if the system is not vented prior to the condensate pump then boiler water backs out and into feed tank on shut down. I have checked the equalizing line (3/4") from the pump to the steam header and it is not clogged. Is it too small? I am assuming the system goes into vacuum as the steam colapses, but what has me baffled is how the water is getting passed the check valves on the feed pumps; I've inspected them and they seem fine.

The second--and more serious--problem has to do with the vacuum itself. The condensate/vacuum pump has never been able maintain a vacuum on its own and judging by the pipe sizes I would assume the system was designed to operate under vacuum (3" risers, 1 1/4" supplies and 12" returns), but perhaps I'm wrong. This year we did some major trap replacement, and not surprisingly, many had clearly failed. There were still some original traps as well. The system as a whole has never really operated well (uneven heat, and under-heated on top floors), so I am looking forward to this season with the new traps. But here's the question: seeing as it is essentially a gravity retun system as is, should I wait to see if it can maintain a vacuum with the new traps (we're still waiting for some repalcement traps, so I haven't fired the thing up yet), or should I go ahead and forget about the vacuum thing altogether and install main and riser vents to enhance performance?

Thanks in advance for your help,

Jim_from_Worcester

Jim_from_Worcester

Corrections

I knew something was wrong as I was writing this; the boiler floods at shutdown, not the other way around. That's why the check valve doen't matter. Still, the question of the equilizing line remains.



The other thing is that the system has 1/2" returns, not 12"!

Jimfrom Worcester

Jamie Hall

Vents

Unless you have a really big vacuum pump, you need to have vents.  Most vacuum systems did; the difference between then and now is that the vents didn't open on a vacuum, and modern vents do -- in the old style, as the boiler cooled the vents stayed closed and the boiler continued to steam down to surprisingly low temperatures.  But my inclination would be to treat this thing as a vapour system pure and simple,  Do you have dry returns which all return to somewhere in the vicinity of the boiler before they turn down to the wet return?  If so, where they turn down is the place for major venting.  If you do do that, though, you also should have "crossover" traps at the ends of the steam mains, going over to the corresponding dry returns.  These are piped up from the main to the inlet of the trap, then from the outlet of the trap over and down to the dry return.  The ends of the steam mains and dry returns are also dripped to a wet return, in most systems.  (There are exceptions).  If you don't have crossover traps, then you assuredly need vents -- like Gorton #2s or more -- at the ends of the steam mains.  Otherwise... the air can't get out and the steam can't get in!



The flooding might be slow wet returns... how is the condensate pump wired?  Is it controlled by boiler water level, or by level in the condensate tank?  It should be the former.  I have to admit I'm not wild about condensate pumps -- I'd much rather see gravity allowed to do its thing.



Maybe a sketch diagram?

Jim_from_Worcester

More info on the system

Hi Jamie,



Thanks very much for the insights. I am, not surprisingly, in need of some clarification. If I understand you correctly the old style vents on these systems used to open at atmosphere or greater and the new style are essentially vacuum breakers, opening at a calibrated vacuum? Otherwise, what's the point of having a vacuum system at all?



You are correct that the system is vented where the dry returns turn into a single return above what little wet return now exists. The return is only "wet" between its lowest point and the minimum condensate level allowed by the floats; it becomes "wetter" as the condensate level rises in the bottom tank until the float tops out and pumps it into the upper tank. Furthermore, the water level is decoupled from the boiler because of the check valves on the feed tank (you can see the wet return for the old Kawanee severed under the new boiler; apparently it was left alone until the new boiler was installed.



I'm not sure if the current vent port is original (I think not), and the vent that was installed was intended for a hot water heater. I have since removed the vent body, so now it is just an open port after a ball valve. I leave this open at all times, which further eliminates the possibility of a vacuum, but it sounds like I should replace it with an appropriate vent/breaker. As a note, the boiler only floods when this ball valve is closed.



I will have to double check the arrangement, but as I recall the only traps on the mains are F&T's at each dripped riser; I do not recall seeing traps tapped at the top of the ends of the mains. All traps, of course, drain into the parallel dry returns.



My first year at the building I noticed the condensate pump was wired so it would run in parallel with the burner. The floats were stuck and the solenoids would not open to allow condensate into the boiler feed tank, so the boiler was running entirely on fresh street water. When the there was a call for heat all of the condensate that had backed up in the returns was flushed into the sump as the pumps kicked in. It would run continuously like this until the thermostat was satisfied. The previous plumber had actually installed another auto fill valve directly on the boiler (there was already one on the feed tank) so it could keep up with steam production!



Anyway, I freed the floats, cleaned the solenoids and rebuilt the check valves. Low and behold, the condensate pump fed the feed tank and not the sump, but only when the burner was firing; otherwise it would still dump hot condensate into the sump at start up. So I rewired the pumps to kick in based on the floats and it now runs completely independent of the boiler. The feed tank, of course, kicks in below a certain level.



It almost sounds like I'd be better off bypassing the condendsate pump entirely and installing Gorton #2's at the ends of the mains and tops of the risers (which definitely never had vents), though I'm not sure how many I would need or what the new piping would have to look like at the top of the risers (I assume they cannot be vented into occupied space). What with the "new" boiler and the associated changes to the returns, it seems like it might be more of a hassle than it's been already to reestablish the system as the original engineers had intended.



I will try and get a diagram posted if I have the time.



Thanks for all your help.



Jim from Worcester

Charlie from wmass

Brief point as I need to read more of your post

The vacuum breaker you removed should remain where it was. It's perpose is to stop the Closing the ball valve is what is allowing the system to flood and that is not the place for an air vent it is the place for what was there, a vacuum breaker. 3/4" at least. The new vents do not allow the formation of a vacuum. In the old days vents for vacuum systems would hold closed under vacuum.

Jim_from_Worcester

Clarification

Hi Charlie,



What I removed was a hot water heater vent, not a vacuum breaker. It vents faster at shut down now and prevents the flooding problem, but of course it eliminates any possibility of a vacuum, which it did anyway. What kind of vacuum breaker/vent should I replace it with? A 3/4" , but how many inHg?



Thanks,



Jim

Jim_from_Worcester

Correction of Clarification

Hi Charlie,



You are indeed right. What I removed was a vacuum breaker (for a hot water tank), and not a vent. The only reason I see for using one of these (if I'm correct in assuming that it vents at atmospheric pressure) and not another kind of vent would have been to prevent the hot condensate from pouring into the basement when the returns backed up because the pump only kicked on with the burner. But it sounds like you would install breakers again and not ordinary vents?



Thanks,



Jim

Charlie from wmass

The system needs vented. Without photos I am

Picturing the system, maybe incorrectly, that a vacuum breaker would go there and a vent would go on the returns before they drop. You may end up needing to make a false water line if the returns are too shallow to prevent crossing of steam. Photos or diagrams would help.

Jim_from_Worcester

Venting Vacuum System

I know one thing: I don't understand this system enough. I certainly was getting crossing of steam before the trap replacements, but hope to have solved this problem. As far as I can tell there is no other way (except for faulty F&T's) for the steam to cross; the boiler feed tank vents to atmosphere and the condensate pump is vented on the top chamber, which feeds the feed tank, vents to atmosphere and accepts the equalizing line with the header. I don't possibly see how the steam could back into the returns via this route.



Am right to assume that if modern vents cannot stay closed in a vacuum that vacuum systems, unless equipped with old style vents, no longer exist? If this is true, then is the reason to maintain the condensate pump to help the condensate through low spots on the returns?



Sorry I keep piling on the questions.



I will take some pictures and post them tomorrow if I get the chance.



Thanks again,



Jim

nicholas bonham-carter

7-story vapor

is the pressure forcing the water into the returns? get a good low-pressure gauge, and a vaporstat to see any difference. a spike in pressure can easily interrupt the trap operation, sometimes permanently

somehow the risers are not getting steam up into the top floor rads until the lower floors have been satisfied. if you had an ir camera, you could perhaps see where the hold up was. some buildings fed the top floor first and then the lower ones to alleviate this problem.

i wonder if this system would work better without the condensate/feed/auto/over-fill pump/tank/headache! these introduce an additional complexity which increases the difficulty of diagnosing waterline problems.

just remind everyone as you work through these problems that the system must have worked well when first installed. the goal is to return it to original specs.--nbc

Jamie Hall

Your best bet...

Now that you have Charles interested in this system would be to get him over to take a look at it.  Money very well spent.



Fundamentally, though, you can start looking at it yourself, and get some clarification.



Modern vents -- e.g. Gortons etc. -- close only if they sense steam.  If they are sensing air, the are open and stay open (they should never sense water, of course).  The objective of the exercise being to let the air out of the pipes, so the steam can get in.  Old style vacuum type vents opened to let air out -- that is, if the pressure inside the pipe was greater than that outside they opened.  They closed on steam, just like a modern vent, but they also closed if the system developed a vacuum, whether there was steam present or not.  There is no reason not to use them -- except that they don't have the venting capacity of modern vents and they are kind of rare.



So... we need modern vents.  If you don't have crossover traps  at the ends of the steam mains -- and it sounds like you don't -- you will need good big vents at the ends of the mains.  The idea being to get the air out of those mains so that steam can get to the risers and radiator runouts.  There is no harm to vents at the tops of the risers, as well, particularly in a building as tall as yours.  The air in the radiators will go out through the radiator traps, along with the condensate that forms in the radiators, and go into the returns.



Any condensate that forms in the dry returns has to get back to the boiler, as does any that forms in the steam mains -- so they must pitch back to the boiler area, where they drop to the short wet return.  It may be (I can't tell without looking) that your f&ts at the ends of the mains serve the purpose of letting condensate over into the returns -- which is fine, but they won't let air by.



I love the way the condensate pump was set up when you started playing with the system -- it is truly amazing what one finds.  That said, I don't like condensate pumps.  They flood boilers.  If there needs to be a pump at all, in my humble opinion it should be controlled by the boiler water level, not by the level in the condensate tank.  And if you are using a condensate tank and a boiler feed pump, and you have an autofill, the autofill should be sensing the level in the condensate tank, not the boiler, and should feed the condensate tank, not the boiler.  But I'd rather let gravity handle the whole thing, if possible.



But as I say, your best bet is to persuade Charles to come look at it...

Jim_from_Worcester

Chapter Four: it begins to sink in

I have double checked and there are no vents anywhere in the system, nor does it appear that there ever were any. The port where the dry returns turn down before the condensate pump is original, but I have no idea what used to be there. Maybe it was the only vent for the entire system. The valve that runs into the erstwhile wet return now remains shut (it would just drain on the floor anyway) and the line that I assume fed into the old vacuum unit now drops into the condensate pump.

 

Now, on inspection of the pump I noticed that the return would never be wet; it enters high on the strainer and the float would top out before the water level could reach the top of the pipe. Also, the equalizing line enters at the top of the inlet for the bottom tank and not the top tank as I had said before. The check valves in this line seem to be installed correctly according to Hoffman’s diagram, so I still don’t think steam would back into the system this way, and I have never seen it do so.

 

Here are some photos.

There is only one horizontal main that takes off from the header and makes an entire loop (with occasional branches to odd risers) around the perimeter of the basement. All of the risers and some independent returns are dripped and trapped, but not one is configured as a vent; they are all handling condensate. Not one riser is vented or has original ports for that purpose. There is something curious about the installation of some of the F&T traps: some are installed below the level of the return and some have an elevated leg from the outlet to the return. Are these proper configurations?  In both cases it would seem that the trap would be constantly open until steam pressure, if great enough, cleared the line and shut the thermostatic valve.

 

I used get heavy wet steam and entering the system and priming just downstream from the boiler, but have unclogged drips and returns and rebuilt the trap where the header enters the main and this has nearly eliminated the problem. Controlling the flooding and skimming the boiler has also helped. It does not appear that pressure is causing any water to be forced into the returns, and with the exception of some low spots here and there all the condensate drains fairly well (if slowly) by gravity alone.

 

When steam is being produced I have always thought (and still do) that the lack of vacuum is causing it to back up behind slowly venting radiators with failed traps. Then, as steam enters the returns, even if slowly, the traps higher in the system would tend to stay closed (unless failed open, which is why I think they worked at all) and heat would tend to move slowly to the upper floors as the air, by hook or by crook, was gradually purged. I think the system used to have one heck of a vacuum pump, as Jamie mentioned it would need in order to function properly.

 

I would love this thing to operate as intended as paying for the installation of all those vents does not really appeal to me. But if I can’t get the thing to produce a vacuum then that might be my only choice. The building might be rehabbed in the coming years and a new hot water system installed, so I have tried to be economical with the improvements. Still, we spend so much every year on fuel that the payback for fairly simple fixes may be less than a single season. The cost of installing vents may be borderline, but I am not sure. Then again (yes guys, it’s beginning to sink in) the “new” pump may be too small anyway, even if it can produce a vacuum, in which case I’d sure be interested in those old style vents.

 

I have included some photos of the condensate pump, turn down of the returns at the old valve for the wet return arrangement.

Jamie Hall

It may be possible

for you to do this thing in stages.  The first stage which I would do is make darn sure that all the radiator traps are working as advertised.  Then you will know that any air which is trying to get out the radiator as the steam tries to get in the other side can do so -- but that steam can't get into the returns that way.



The next step is to make sure that the air can get out of the system.  You have an opening where the dry returns (which pick up the returns from the radiators, yes?) turn down.  It is quite possible that there was a vacuum pump there at one time, but it's not needed.  What is needed are a two or three Gorton #2s on an antler set up.  Now the air can get out of the system.



Now to improve the evenness of heat, you need some way to vent that steam main.  But I would hold off on that until you do steps one and two above.



Perhaps the most important thing at this point, though, is to make sure that there is a clear path for the air from every radiator's return to the vents you are going to install -- there should be one and only one trap between each radiator and the main vents, and that is the trap on the outlet of the radiator.



The f&t's you mention which go up and over and back down from the steam main to the dry return sound as though they were crossover traps once upon a time, but an f&t won't work that way.  It might be possible to replace them with big radiator type traps -- 3/4 inch -- and improve things somewhat in terms of evenness.



I really, seriously, can't recommend putting a vacuum pump back on.  They are a pain in the neck from the maintenance standpoint, and are pretty expensive, and if the system is adequately vented they won't get you anything.



I have neglected to ask -- what pressure are you running?  I would suggest trying it at a nice low pressure -- no more than a couple of pounds, max., and dialing it down from there to see how low you can go.  It can't hurt, and just might help.

Jim_from_Worcester

Now here's my plan . . .

Thanks Jamie,



I agree with you on the traps; they all have to be working properly before I can move to the next step. I am awaiting a new shipment of cage units to replace the wrong ones which were improperly installed ages ago (and nearly all failed). I have shut the supply valves on these units, but I'm still worried about steam entering the returns because some of the supply valves are thermostatic, which I find never completely close.



I like the idea of replacing the strange F&T's with 3/4" thermostatic traps (can they handle that much condensate--14 radiators worth?). Otherwise I will attempt to reconfigure them so they are operating properly. As a note, I don't think these were crossover traps as you describe because both the traps and piping seem to be of a more contemporary vintage. In any event I will fire the boiler tomorrow and test to see if they are functioning properly and see if I can detect how far the condensate is backing up in the main and riser.



At this point I will test to see if the system can maintain a vacuum, (why not?) but will likely reopen that vent port if I cannot. There is a check valve here that only will allow airflow into the return. There is a vacuum breaker on the strainer basket inlet cover (you can just see it in one of the photos) and the equalizing line has a check valve prior to the vent on the top tank that flows toward atmosphere, so it should act as vent for the returns and be able handle a vacuum. So I like the idea of flipping the check valve in the return line before the pump and installing a vent as you suggest. I just have to address Charles' concern and make sure that vacuum breaker is set to open correctly.



At this point, if all is working and I can get the upper radiators to heat in a timely manner, I will leave well enough alone. If not, I will splurge and have some Gorton #2's installed on the main just beyond the points of the furthest two risers (based on Jamie's earlier comment I found out about Hoffman #76's. Are these still available anywhere?). If this still doesn't solve the problem, then riser vents enter the picture.



Jamie, I completely agree about the vacuum unit; It's been such a pain in the arse so far I'd love to see it go away. But it's there, and if I can produce vacuum without too much more effort then I will live with it for a while. If I can't, then I will start loading on the vents.



The system generally runs at 2 lbs, but it has been known to climb as high as 5 on very cold days.



As a note, I use Lyon's Plumbing here in town and these guys know their steam. Together we have often puzzled over this system and they have been accomplices in my improvements. They don't however have the time to get to see all the intricacies of the system as I do, so they rely on me to tell them where to focus. Invariably they come back with new insights. Most of the work I've had them do is in the boiler room proper. I would love to have Charles come look at what is here, and perhaps that may be in the future (Charles, how can I get in touch with you?). But I will certainly be using Lyon's going forward and running this plan by them.



Thanks all of you for your help. I am truly gratetful to have a forum like this with such smart folk at my disposal!



I will let you know how I make out (or what I screw up).



Jim

Rod

You really need a Steam Pro !

Hi Jim- I've been following this post and I most emphatically agree with Jamie's recommendation that you need to get hold of Charles Garrity. Here's a link to his page in the "Find a Contractor" section.

http://www.heatinghelp.com/professional/221/Charles-Garrity-and-Son-Plumbing-and-Heating

Charles is a very experienced steam pro and that appears exactly what you need.  From your description I'm not overly impressed with people you presently have working on your system as there are obvious problems that any "pro", who really understood steam, would have immediately recognized.  Having a system, that should be running on ounces of steam pressure, run as high as 5 PSI is absurd and is a huge waste of fuel.

Whether you want to do the work yourself or have the local guys do it isn't important. What is important is having a definite viable plan for fixing your system  and that's where you need a steam expert.  Vacuum systems are the "Cadillac" of the steam systems and once you get the "mickey mouse" work that people have done to it over the years straightened out, you will have a very comfortable and economical heating system.

- Rod

nicholas bonham-carter

7 story over-pressure

If I were tackling this, I would get a vaporstat on the boiler to get the pressure down to 12 ounces or so verified by a good low-pressure gauge on the same pigtail.

Many traps and vents will not work at pressures higher than 3 psi, and that may explain all the trap problems.

The fact that the pressure has been set so high is no doubt an ineffective cure for short-cycling as the boiler, perhaps over-sized tries to force the air out.

For every ounce of pressure above atmospheric, there is an increase in the level of condensate of 1.75 inches, and so you can see that column of water higher than we can reach just waiting for boiler shutdown to suddenly fall.into the condensate tank!

If this is a big boiler, then it should have a "hi-lo" burner on it which would adjust the firing rate to the load by means of pressure. As the vaporstat cuts out at say 4 ounces, the burner switches to "lo" fire to maintain the pressure at the lower pressure alternating between lo and hi. Often this feature has been ignored by installers (even those who "know their steam") who are ignorant of it's purpose.

Definitely get the real expert on this project to get this system back to it's original state of operation. I am going to guess that char lies services will be paid out of fuel savings in a couple of months.--NBC

Jim_from_Worcester

Pressure level

Thanks for the information. Whatever issues I'm having do not reflect on my plumbers; they are extremely competent and without them (and Dan Holohan) I would have little or no understanding of the system at all. Whatever I have posted here that seems off arises either from my misunderstanding or reporting or previous mistakes by other plumbers who didn't understand the system.



That said, I will be placing a call in to Charles. The pressuretrol is something I've never touched, and the low pressure of the system is something I've never had a very good handle on; I keep my ears pricked up for mid-cycle troubles and turn the heat down at the first sign of hammering (the system is almost always at 2 lbs or less; 5 lbs is a rare occurrence). Thank you for the description of the hi-lo switch, I had no idea what that was for. It certainly sounds like many of my problems are stemming from the settings on these controls. I will post them for the curious.



Jim

Jim_from_Worcester

Oops

Dave in QCA

Vacuum = Absence of Air

When a Vacuum system is operating correctly, it will enable the boiler to produce steam well below 212 F, and at the beginning of a firing sequence, the steam will go speeding throughout the steam mains and risers to all of the radiators, because their is no air to cause an impediment.  When the system is not working correctly, performance will be very poor!



Some points of advice:

1.  Get a steam pro involved who has experience and knowledge of vacuum heat systems.

2.  Traps must all be working correctly, or the system will not function.

3.  The system may or may not have had vents, as many systems used crossover traps to allow the steam mains to vent to the condensate return piping.  

4.  If there are vents, they must be vacuum type, or there will be no way to pull a vacuum on the system.

5.  If the system was originally engineered and designed as a vacuum system, it really should be maintained that way.  In most cases, vacuum design enable the use of smaller pipe sizes throughout.  Switching to standard low pressure steam operating can cause problems because of inadequate piping sizes.



Here is a link to a Hoffman Specialty publication.  http://completewatersystems.com/wp-content/uploads/2011/05/7250.pdf   Page 6 shows vacuum systems.



Keep us posted with your progress.  We don't see many vacuum systems on the wall here.  If you report your progress, we can all learn a thing or two.



I'd love to see more pictures of the system!  Boiler, radiators, piping, and of course.... how about the building too!

Jim_from_Worcester

Vacuum System

Hi Dave,

Thanks for the input. I would love to see how the system works under vacuum. Today I tried the pump without firing the boiler and no luck. I am going to isolate the vacumm unit tomorrow and see if I can get the vacuum switches to open (their calibrated to do so at 8 inHg). The I will know if I have leaky piping or a bad unit. Solenoids and check valves may not be up to snuff either.

1. I put a call in to Charles. Hoping for a pow-wow with my plumbers.

2. I working to get all the traps working

3. Here's where my idiocy really shines: all that talk of crossover traps had me imagining things; I was picturing seperate mains connected though a single trap that only handled air and completely forgot that F&T's handled air. I have crossover traps at all the dripped risers, and, with the exception of the oddly piped ones, they are all venting to the returns! I really feel stupid on that one. It's a good thing I don't do this for a living. How many posts did it take before that sunk in? So, all my mains, theoretically, are vented through the F&T's at the risers.

4. If additional vents are needed, would the Hoffman 76's work?

5. My inclination is to restore the functionality of the system, because it's what is there, but it has been tinckered with (and neglected) so much and with equipment that I don't know is appropriate. I'm not even sure if the current vacuum unit ever worked properly. So, short of sinking a fortune into it, I may be forced to optimize its last years as a vapour only system.

I will try and send additional photos if I have the chance.

By the way, here are settings for the controls:

Operating Pressuretrol (mounted above fill valve on pig tail): 3 lbs cut-out with a subtractive difference of 0

Hi limit Pressuretrol (mounted on antler configuration anove boiler, on left with pig tail): 10 lbs

Hi-lo switch (mounted on right of antler w/ pig tail-- the gauge is in the middle of antler): looks like 5 lbs.

Thanks again,



Jim