Return a pumped system to gravity feed hartford loop?

GBC_illinois

I have a steam boiler in a share building which was gravity feed once, but when the boiler was replaced in 2011 (before my time), a Shipco condensate tank and boiler feed pump was installed. Anyway, the pump has caused problems and I'm keen to get rid of it now because I think the condensate comes back fast enough for gravity feed to be viable. I don't think this was even tested when they ripped out the gravity feed system. The system we have is a Weil-Mclain LGB-9 steam boiler. It's a 2-pipe system in a shared-ownership coop building in Champaign, IL.

Many of the boiler specialists I've spoken to locally do not seem to understand what a hartford loop is or the theory behind steam heating. So, I'm determined to oversee this change myself. I've read about the hartford loop in the Holohan book and have seen the diagrams, but I have some particular questions about best practices:

1. According to my calculations, during normal operation, there is a maximum of 7-10 gallons of water out in the pipes at any given time when the steam boiler is running, usually much less than that.
   
2. Right now, the boiler feed pump and ShipCo tank keep the boiler's water line at a fairly exact level, but with gravity feed, I assume this water level would vary now based on the amount of water out in the lines. If the 10 gallons of water I measured was out in the lines, that equals a 3.5 inch difference on the sight glass of the boiler. Is that ok on gravity-fed systems for the boiler's water line to vary by 3.5 inches during operation? I would not want to use make-up water during normal operation; I would want the water line to be near the low-water cutoff when the max water is out in the pipes, and then when the condensate all returns after the boiler cycle, for the water line to be 3.5" higher. Is this reasonable?
   
3. If 3.5" of water line variation is not OK, does that mean going back to gravity feed is not possible? Or is there another way to accommodate it? I've heard of using a reservoir tank, but I would want it to be completely gravity fed, not pumped, or it would defeat the purpose of this change.
   
4. I assume that a new type of make-up water feed system would be required, but I can't find very much information about this. Can someone point me in the direction of a diagram or devices that control make-up water for gravity-fed systems, and how to set the water level at which make-up water is allowed in? I would want this to be not much above the low-water cutoff.
   
5. Would I still keep the F&T steam traps at the end of the mains & condensate lines if I return to gravity feed? My building has no radiator steam traps, which I know is odd. But trust me, I've looked even for the pinhole Richardson-style elbow traps, there are just none of them. Steam goes right back into the condensate lines in the basement, trapped only by the end-of-line F&Ts. Do I get rid of them or keep them if I return to gravity feed?
   
6. When they say to put the hartford loop connection 4" below the water line in a commercial size steam boiler, does that mean 4" below the lowest water line (like the low-water cutoff?). I don't hear a lot of talk about boilers where the water line varies in height during operation. It looks to me like I could use the existing connection from the pump into the equalizer / header drip line, which is 26" off the floor (see photos). The boiler water line is currently 32" but in a gravity feed scenario, I was thinking I would lower the water line to about 30" before the make-up water was triggered? and keep the low-water emergency cutoff where it is now at about 28.5-29"?
   

I appreciate any help you can provide!

Mad Dog_2

I've seen that happen often where something was "engineered" and "In the spec" whether it was needed or not. Why not do a little repiping and run system without it.  You could quickly put it back. Some great steam minds may disagree, but in these situations I have made a Mad Dog 🐕  Trademarked (kidding) Bazooka Bottle Return reservation out of 8' or 10" black pipe.  It adds an extra Reservoir condensate reserve to assure boiler and system have time to catch up with eachother.   With drains on both ends it also serves as a good mud-puppy Trap that yiu can easily power flush out with a hose. Mad 🐕  Dog

Long Beach Ed

Look for the height of the lowest steam carrying pipes. If you remove the traps, you will need to ascertain that each return ends in a wet return. Otherwise steam will "short circuit" back into other returns and circulation will stop. That "return manifold" looks like it may be a bit high, as the original boiler may have had a higher waterline. Measure it.

The range of the operating waterline should be in WM's installation instructions.

DanHolohan

May we see a photo of a typical radiator? Please show a closeup of the supply valve and another of the return. And please show us the whole radiator as well. Thanks.

Jamie Hall

This is not only feasible, but may be the best thing to do. However, there are some complications.

Before the comments, though, is this really a two pipe system? No vents on radiators, and the air and condensate leaving the radiation from an outlet positioned low at one end? If it is, the circulation and function of the pipes is quite different from a one pipe system!

First, as @Long Beach Ed mentioned, check the elevations of the piping. Assuming it is a two pipe system, you have three kinds: steam mains, dry returns (or they should be dry -- I'll get to that), and wet returns. It is important that the steam mains and the dry returns be well above the water level in the boiler -- 28 inches for every pound of pressure you are operating at (I'll get to that, too). But it is equally important that the wet returns and the connections of drips from the steam mains to the wet returns be below the operating water line in the boiler. Sometimes when a new boiler is installed, the operating water line is significantly lower than the old one unless the installer was paying attention, and this can lead to a good deal of trouble.

Second, there is never supposed to be steam -- or steam pressure -- in the dry high level returns in a two pipe system. If this really is a two pipe system, there has to be some way to keep steam out of the high level returns -- otherwise you will have steam trying to get into the radiation from both ends, and no way for air to get out, and the heating will be poor. So -- aggain the question here is, is this really a two pipe system? If it is, the steam may have been intended to stay out of the dry returns by limiting the amount that can get into the radiation by means of calibrated valves or orifices, rather than relying on traps. If this is the case, then you will need to arrange your boiler controls to maintain a very low pressure -- cutoff measured in ounces per square inch.

Third, again if this is a two pipe system, the question arises as to how does air get out of the steam mains. In some there were vents at the ends of the mains, but in many -- perhaps the majority -- there were crossover traps which allowed air, but not steam, to go over to the dry returns, which in turn were vented where they joined at the boiler before they dropped (note that if this is one pipe, the "returns" must be individually vented and drop before they join; if this is two pipe, the returns must be joined before the drop and vented together).

On the question of water out in the system as steam. I'd recheck that calculation. It is unlikely that more than 10 minutes worth of steam at the firing rate of the boiler will be out in the system at any one time. but yes, it is more or less normal for the water level to drop some when the boiler is running.

I may have more thoughts as the conversation here progresses!

ethicalpaul

7-10 gallons, are you sure?

that would be 12,000 - 17,000 gallons of steam

JUGHNE

You mentioned "shared building"....

Just to be clear, are there any motorized zone valves?

Only 1 tstat/control for entire building?

EBEBRATT-Ed

I would not remove the condensate tank until you are sure the system will work on gravity. Just pipe the return to the condensate return tank with a tee and pipe it to the boiler with a valve in it and also a valve to shut the return to the receiver. Basically "pipe around" the receiver. The water make up can be piped to the boiler.

That way you can switch back and forth to test things out.

Some of the newer boiler hold less water so a surge tank like @Mad Dog_2 mentioned is needed,

As the boiler steam water is constantly returned at the same rate the boiler is steaming so the boiler water level should only drop a little unless your returns are plugged up.

Pictures of the radiation and steam drip will show weather traps can be removed or not

GBC_illinois

DanHolohan said:

May we see a photo of a typical radiator? Please show a closeup of the supply valve and another of the return. And please show us the whole radiator as well. Thanks.

I really appreciate you taking the time to look into my situation! Here's two radiators: the carpeted room is a residential one; they all use Danfoss RA2000 thermostatic radiator operators on the valves. The other one is in a public space and I've removed the RA2000 to ensure it always works at maximum.

DanHolohan

Thanks. Are there any radiators with the original supply valves?

GBC_illinois

ethicalpaul said:

7-10 gallons, are you sure?

that would be 12,000 - 17,000 gallons of steam

This is a 15-unit coop building with between 60 and 70 radiators, and its clear this boiler is oversized for that load. It is possible that some of this 10 gallons is leaving the system somehow, although I'm not aware of any leaks. I came to this "10 gallon" number via two different methods:

Method 1) I have a water meter attached to the makeup water line going into my condensate tank. The tank has a sightglass and the makeup water line will refill the tank to a level that is about 2.5" below the top of the sightglass, which allows about 16 gallons of empty space (or extra water coming back from the return). I drained this tank to just the level where the makeup water will quit refilling it, at the 2.5" below top mark. I noted the water meter reading. Then, I let the boiler run for a few heating cycles (we have an indoor thermostat currently, so it cycles on and off to maintain). I looked back at the water meter, and it had refilled the tank with an additional 9.0 gallons since my initial read.

Method 2) I shut off the makeup water valve and drained the condensate tank to half, so the middle of the sightglass was at the center of the cylindrical tank. The tank is 22" diameter by 30" long, which means at the center of the cylinder it is 660 cubic inches, or 2.85 gallons of water, per inch of rise on the sightglass. With the makeup water turned off, I can watch the sightglass go up and down with the boiler feed pump triggering and the condensate returning. I used my phone to capture a 2 hour time lapse of this with a ruler taped to the sightglass, and saw about 3.5 inches of difference between the low and the high of the cycle. 3.5" at 2.85 gallons per inch = ~10 gallons.

However, as I said, it is possible we are losing water somewhere in the system, because I have noticed in the coldest periods, I'm using more makeup water when there doesn't seem to be any way for water to be leaving the system except for the vent pipe on top of the tank (and I actually measure the condensed steam that comes out of that -- it isn't much!)

GBC_illinois

JUGHNE said:

You mentioned "shared building"....

Just to be clear, are there any motorized zone valves?

Only 1 tstat/control for entire building?

No motorized zone valves, and one thermostat for the whole building (an indoor RANCO thermostat which I'm hoping to replace with a heat-timer eventually).

GBC_illinois

DanHolohan said:

Thanks. Are there any radiators with the original supply valves?

Not that I know of. Each unit in this building has been owned by at least 7 people since 1920 when it was built, and each of those owners made changes. I have found a few radiators with different valves than the RA2000, but none that weren't thermostatic.

You might find it interesting that this building was originally heated by steam piped in from a steam plant several blocks away. I learned this when I found a disused steam main headed out of the building, and then researched and discovered that there was a steam plant in downtown Champaign, blocks away. I suspect that this "city steam" ended no later than the 1950's or so, and which point I wonder if the radiators had Richardson-style pinhole elbow steam traps, but so far I haven't found any in the radiators which had to be repiped.

You might also find it interesting that when they replaced the boiler in 2011, one of our boiler contractors (can't figure out who, there were too many with their hands on it) decided they should set it to 7psi of pressure, which caused terrible, thunderous water hammer (my dad said I should sue the seller for not disclosing). Since I bought in and took over boiler operation, I have since learned that it can operate just fine on 13oz of pressure -- 1/10 of what it was! And now the water hammer is gone. I suspect it could operate lower than 13oz and still heat every radiator, but the boiler is so oversized that it needs to cycle too rapidly for my taste to try any lower.

I wish all those supposed boiler pros which were here charging us 10's of thousands of dollars to service our system in the last few years had read your book! It could have saved us so many headaches!

DanHolohan

I'm pretty certain the original supply valves had a variable orifice in them to limit the amount of steam that could enter the radiator at low pressure. The Dead Men often deliberately oversized their radiators so the orifices could do their job. No steam entered the return because the radiator had the ability to condense all that could squeeze through the orifice. I think the Danfoss valves are now acting as variable orifices. That's what's allowing it to work without traps.

GBC_illinois

Jamie Hall said:

This is not only feasible, but may be the best thing to do. However, there are some complications.

Before the comments, though, is this really a two pipe system? No vents on radiators, and the air and condensate leaving the radiation from an outlet positioned low at one end? If it is, the circulation and function of the pipes is quite different from a one pipe system!

First, as @Long Beach Ed mentioned, check the elevations of the piping. Assuming it is a two pipe system, you have three kinds: steam mains, dry returns (or they should be dry -- I'll get to that), and wet returns. It is important that the steam mains and the dry returns be well above the water level in the boiler -- 28 inches for every pound of pressure you are operating at (I'll get to that, too). But it is equally important that the wet returns and the connections of drips from the steam mains to the wet returns be below the operating water line in the boiler. Sometimes when a new boiler is installed, the operating water line is significantly lower than the old one unless the installer was paying attention, and this can lead to a good deal of trouble.

Second, there is never supposed to be steam -- or steam pressure -- in the dry high level returns in a two pipe system. If this really is a two pipe system, there has to be some way to keep steam out of the high level returns -- otherwise you will have steam trying to get into the radiation from both ends, and no way for air to get out, and the heating will be poor. So -- again the question here is, is this really a two pipe system? If it is, the steam may have been intended to stay out of the dry returns by limiting the amount that can get into the radiation by means of calibrated valves or orifices, rather than relying on traps. If this is the case, then you will need to arrange your boiler controls to maintain a very low pressure -- cutoff measured in ounces per square inch.

Third, again if this is a two pipe system, the question arises as to how does air get out of the steam mains. In some there were vents at the ends of the mains, but in many -- perhaps the majority -- there were crossover traps which allowed air, but not steam, to go over to the dry returns, which in turn were vented where they joined at the boiler before they dropped (note that if this is one pipe, the "returns" must be individually vented and drop before they join; if this is two pipe, the returns must be joined before the drop and vented together).

On the question of water out in the system as steam. I'd recheck that calculation. It is unlikely that more than 10 minutes worth of steam at the firing rate of the boiler will be out in the system at any one time. but yes, it is more or less normal for the water level to drop some when the boiler is running.

I may have more thoughts as the conversation here progresses!

WOW thank you so much for this detailed analysis and great questions!

It is definitely 2 pipe -- I've attached some pics below of the basement piping where you can see the separate, uninsulated condensate line running alongside the insulated steam main. There are two separate loops of steam & condensate mains in the basement, one running north from the boiler and the other running south. I've shown both.

This system currently runs at 13oz of pressure.

As far as I can tell, the "dry" portion of both the steam main and the condensate line is all well above the boiler water line. Of the two loops of basement boiler piping, the north loops are about 76" high in the "dry" portion, and the south loops end about 72" high in the "dry" portion. After that, each of these four pipes goes through an F&T steam trap before entering a condensate collection pipe which is about 25" high (pictured) before it goes to the condensate tank. The boiler water line is currently 32".

RE: "Should never be steam in a dry return, otherwise you have steam coming in from both ends of radiator" - THIS DOES HAPPEN. Because our boiler cycles with an indoor thermostat and runs on only 13oz of pressure, it typically doesn't run long enough for it to be a huge problem, but some people do experience a bit of warmth on their radiators even when their control knobs are off because there is steam in the returns. The basement condensate pipes get to about 200 degrees, as well. We have no radiators in the basement, so we don't mind the uninsulated condensate pipes acting as a linear radiator. I know this arrangement is not typical, but as far as I can tell this setup is not causing a problem at our low pressure and non-constant usage.

RE: Venting: see the pics I attached. There are only two air vents in the entire system, one on the last foot of the dry return of each condensate line as it returns to the boiler, as you noted.

RE: Water calculation: I replied to another user in this thread about how I calculated that. I did it in two ways, and they both came out similar. That being said, I do wonder where the water goes, because the return rate does not seem very consistent. There are times where there is more variation than others, and I wonder if maybe whether the individual radiator controls affect it? Like, if somebody goes around and shuts all the radiators in their unit off, do they hold onto water and then release it later when they turn them on again? This subject has confused me for sure.

Here are the pics, and thanks again for such detailed analysis and insight!

GBC_illinois

DanHolohan said:

I'm pretty certain the original supply valves had a variable orifice in them to limit the amount of steam that could enter the radiator at low pressure. The Dead Men often deliberately oversized their radiators so the orifices could do their job. No steam entered the return because the radiator had the ability to condense all that could squeeze through the orifice. I think the Danfoss valves are now acting as variable orifices. That's what's allowing it to work without traps.

That makes a lot of sense!

What do you think about the boiler water line varying by 3.5" during operation? This is what I suspect will happen if I return to gravity feed without some other intervention. Is that too much variation? It's been a while since i read your book but here are the measurements I've taken so far:

Boiler water line is currently 32" with the pump, but I think it could be minimum 30" with gravity return, with makeup water entering once it gets below 30".

The bottom of the steam header is 70" high.

If I'm right about how much condensate collects in the piping during operation (10 gallons max) then the boiler water line would vary about 3.5", but just to be safe let's say that it would vary by 5", making the water line between 30"-35" high from the ground, which still gives another 35" to the bottom of the header.

It looks to me like the close nipple of the hartford loop can go into the equalizer at 26" high, for which I can use the existing pipe T from the pump. So that would give 4" from the nipple to the bottom of the water line.

Thank you again for all you do! Your book was priceless.

DanHolohan

Thanks. I would try the loop, bypassing the pump, but keep the pump piped off in place until we see how the loop does on its own.  I think this will work without the pump, but there’s only one way to find out. 

Long Beach Ed

The condensate return rate will vary with heating conditions. Bear in mind that there are common low water cut-off controls that will shut the boiler and wait a settable period of minutes to permit condensate to return in extreme conditions. They won't feed fresh make-up water until that delay is passed. So if the system is slow to return condensate on an occasional basis, as with a severe temperature drop, the proper control can keep from flooding the boiler with new make-up water.

Jamie Hall

Agreed with @DanHolohan -- which is no surprise. If you do find that you need more water storage at the water line, but not pumping, you can repurpose that receiver at a slightly higher level to do that. Connect its vent to the equalizer and there you are.

Sounds as though no one really mangled your system in the past. Good thing. I'm not convinced that you will need the F&T traps at all, once you get it back to equalized gravity operation.

WMno57

@Fdarby82 I'd like to introduce you to @The Steam Whisperer. You guys are less than 2 hours apart on I-57.

JUGHNE

This little boiler holds 12.4 gallons of water. The receiver/accumulator holds an additional 3.28 gallons.

There is not that much EDR connected, but there is 147' of main before it drops down to the wet return on the floor. The wet return is only 1" copper--adequate for the load.

This added tank was made from the remains of the 6" header pipe of the old boiler removed. A welder built the tank with fittings etc.
It is not considered a "pressure vessel" (subject to all that that would involve), rather is is simply more sch 40 pipe in the system.

It simply gives you a boiler with more water capacity, of which this tank does not have a fire under it.

This and another has been reviewed and passed by the State Boiler Division.



The picture has everything skewed somewhat, I don't put things in this crooked.

The end center of the tank is at the water line.
There is a steam equalizer connected to the top of the tank.
The bottom is connected to boiler return inlet.
There are plugged ports for cleaning and flushing, top and bottom.

You can see the equalizer connection with a close nipple.

This is a small example of what you could do if your return is too slow.
So IIWM, I would provide the ports for future equalizer and return.

BTW, I did replace the CycleGard with a standard LWCO.

Long Beach Ed

Beautiful work!

GBC_illinois

One question that I don't think anyone has addressed yet: Is it unusual for a boiler's water line to vary by 3.5" during operation? The only boiler I've ever worked with is my own, and I always hear references to the boiler's "waterline" as if that is a fixed place on the sightglass. Is 3.5" of variation OK?

Jamie Hall

It's more than I'd like -- but Cedric's drops almost as much.

Long Beach Ed

Check the manufacturer's data for the operating range of the waterline. If it's not in their material, you can call them. I agree with Bro. Jamie... it doesn't sound unreasonable.

JUGHNE

With the pump feeder control you have did you have that 3.5" variation in the water level?

From my experience with those M&M's, those controls usually had very little differential from pump on to off.

Mad Dog_2

That Tank is neato.  We designed had one made for a job. Ill try to find a picture of it.  Mad Dog

Long Beach Ed

JUGHNE said:

With the pump feeder control you have did you have that 3.5" variation in the water level?

From my experience with those M&M's, those controls usually had very little differential from pump on to off.

I believe the 193 control has 3-1/2 inches between on and off.

JUGHNE

It looked like a 150 to me.

GBC_illinois

JUGHNE said:

With the pump feeder control you have did you have that 3.5" variation in the water level? From my experience with those M&M's, those controls usually had very little differential from pump on to off.

With the current tank and feeder pump, there is about an inch at most of variation. Probably more like 3/4 inch. 

JUGHNE

This is another receiver/accumulator tank.
It was done years before I got there. Took me a long time to figure what it did.
The bottom of the tank has 2 taps into both boiler return ports
There is an equalizer line on the tank top out of the side of the header .
The water fill goes both into the tank and the boiler.

This may have been over kill for this system. Total refill to water line requires 95 gallons.

3" steam mains off the header go about 110' to each EOM, 2" return that distance for each main to wet return at boiler. 1238 total connected EDR.

No true equalizer drip, just 2 4" boiler risers.

System runs from 4 oz to 12 oz, quiet with no hammer except one badly piped last rad.




This is the back end of the tank. The gravity return Hartford Loop connection is below the water line of the tank. Inspectors have taken pictures of this, one said... "and they said that this would never work".
This included numerous insurance inspectors and the State Boiler Inspector, (The Big Guy himself).

The 2 EOM returns have since been separated and each drop down to the wet return .



The front showing water supply connections.

I even added a blow down for the manual reset LWCO probe boiler port, after inspection required 2nd LWCO.

The tank is a real sludge collector for keeping the boiler cleaner.