Vacuum for one pipe Paul air line system

SWEI

Nothing smaller than 1 HP?

I'd love to find something smaller in 3Ø (say 1/2 HP), run it on a VFD and modulate the vacuum.

jumper

I suggest a retail 110V compressor is all you need if you seal the air out.

nicholas bonham-carter

Someone in the last few years discussed using a shopvac as the vacuum pump. Would that work?--NBC

jumper

More off the top of my head. An ejector moves air in proportion to its motive fluid mass wise. So when it pulls a vacuum with say one SCFM motive air it can evacuate more ACFM as vacuum increases. A positive displacement pump (like liquid ring) evacuates roughly same amount of ACFM regardless of pressure.That means less SCFM as vacuum increases.

Pumpguy

@SWEI, So far, 3/4 HP liquid ring vacuum pump is smallest I've found. Unfortunately they're more expensive than the 1 hp pump I currently use. I can get the price down to an acceptable (hopefully) level using salvaged/reconditioned parts along with a new motor, shaft seal and gasket.

Back in the day, Nash Engineering Co, (my former employer) made smaller ones, but not anymore.

I can build these pumps with either 3 phase or single phase motors.

I have been toying with the idea of converting a 1 hp 14 ACFM pump by shortening the impeller and adding a filler block in the housing. This could reduce down to as low as 1/3 hp, 4 ACFM. If done, this would be a one off, toolroom model, and replacement impeller and housing would not be readily available.

Keep in mind that the bigger pump would run for a shorter time to cycle between the settings of the vacuum switch. The smaller pump would have to run longer. The OFF, or leak down time would be the same, no matter which pump is used. This of course depends on the air tightness of the system.

@NBC, A typical shop vac is a centrifugal type machine, powered by a brush type universal motor. When dead end connected to a vacuum gauge, max vacuum I have seen is only around 8" Hg.

That, combined with brush and commutator wear, small bearings, and most importantly the NOISE, makes them, IMO, not suitable for anything more than experimental use.

@jumper, You are correct. The performance curve for a typical liquid ring vacuum pump is pretty flat, in terms of ACFM, as it goes out from atmosphere to operating vacuum. As the air is discharged, it compresses back to atmospheric pressure, so the SCFM drops off as the pump pulls out to the required system vacuum.

Pumpguy

@SWEI, Liquid ring vacuum pumps are speed sensitive. The blade tip speed is critical. In this case, the motor is 3450 RPM. As a result, they don't typically lend themselves to speed variations controlled by a VFD. Speed changes can cause the liquid ring to collapse, causing the pump to go into a stall.

SWEI

Thanks for the info -- that does make sense.

I'm wondering if a small lab/bench type vacuum pump might be usable for residential system sizes. We have monthly auctions (thanks to the the two national labs here) where all manner of interesting stuff tends to show up. Exotic gear that may have cost $5-20k new can sometimes be had for less than $100. Is 4-10 ACFM a reasonable target range to look for?

jumper

An accumulator is a very good idea for a dry pump because it won't like sucking liquid. The so called rotary piston (Kinney,BeachRuss,Stokes) are very rugged.

PMJ

Just a few thoughts for those with one pipe systems considering this. While significant vacuum induced mechanically with a pump makes good sense, simply the achievement of routing all the rad vents with some sort of tubing to a central location would open the door to all kinds of possibilities even without a pump. System balancing now done with different fixed vent rates at each rad would now become dramatically more flexible by simply opening and closing each vent tube for different lengths of time relative to each other at a central manifold. Also possible would be turning off the heat completely in various rooms on some daily schedule. And if all vents close when the burner goes off there is significant natural vacuum to take advantage of. We can all appreciate that very tiny differences in pressures within a system cause the steam to flow in different directions. I have observed in my two pipe in natural vacuum that when the burner is off the vacuum forms slightly faster in rads where the steam is condensing faster and pulls more from the mains on the off cycle than the others. I turns out that this is exactly what I want and has a significant impact on evening out the heat. I believe it would be quite the same in a one pipe system with all the vents closed on the off cycle. Bear in mind that with traditional 2 way rad vents the rads condensing the fastest simply pull air faster from the room on the off cycle - the opposite of what you really want them to do.

Obviously controlling this central venting manifold would require a control on the PLC level and is a steam enthusiast kind of project. But the cost of that would be very small in relation to the cost and trouble to install the vent plumbing. And the knowledge required to program the simple PLC really isn't much. If I had a one pipe system I don't think I could keep myself from attempting this - I believe the end result would be better than a 2 pipe system. I would do the control before adding the pump. I still haven't added a pump to my system because just the natural vacuum was so effective - and totally free with no maintenance.

ChrisJ

PMJ said:

Just a few thoughts for those with one pipe systems considering this. While significant vacuum induced mechanically with a pump makes good sense, simply the achievement of routing all the rad vents with some sort of tubing to a central location would open the door to all kinds of possibilities even without a pump. System balancing now done with different fixed vent rates at each rad would now become dramatically more flexible by simply opening and closing each vent tube for different lengths of time relative to each other at a central manifold. Also possible would be turning off the heat completely in various rooms on some daily schedule. And if all vents close when the burner goes off there is significant natural vacuum to take advantage of. We can all appreciate that very tiny differences in pressures within a system cause the steam to flow in different directions. I have observed in my two pipe in natural vacuum that when the burner is off the vacuum forms slightly faster in rads where the steam is condensing faster and pulls more from the mains on the off cycle than the others. I turns out that this is exactly what I want and has a significant impact on evening out the heat. I believe it would be quite the same in a one pipe system with all the vents closed on the off cycle. Bear in mind that with traditional 2 way rad vents the rads condensing the fastest simply pull air faster from the room on the off cycle - the opposite of what you really want them to do.

Obviously controlling this central venting manifold would require a control on the PLC level and is a steam enthusiast kind of project. But the cost of that would be very small in relation to the cost and trouble to install the vent plumbing. And the knowledge required to program the simple PLC really isn't much. If I had a one pipe system I don't think I could keep myself from attempting this - I believe the end result would be better than a 2 pipe system. I would do the control before adding the pump. I still haven't added a pump to my system because just the natural vacuum was so effective - and totally free with no maintenance.

I thought the reason for a pump was to achieve steam temperatures in the 140-160F range on demand?

PMJ

ChrisJ said:

PMJ said:

Just a few thoughts for those with one pipe systems considering this. While significant vacuum induced mechanically with a pump makes good sense, simply the achievement of routing all the rad vents with some sort of tubing to a central location would open the door to all kinds of possibilities even without a pump. System balancing now done with different fixed vent rates at each rad would now become dramatically more flexible by simply opening and closing each vent tube for different lengths of time relative to each other at a central manifold. Also possible would be turning off the heat completely in various rooms on some daily schedule. And if all vents close when the burner goes off there is significant natural vacuum to take advantage of. We can all appreciate that very tiny differences in pressures within a system cause the steam to flow in different directions. I have observed in my two pipe in natural vacuum that when the burner is off the vacuum forms slightly faster in rads where the steam is condensing faster and pulls more from the mains on the off cycle than the others. I turns out that this is exactly what I want and has a significant impact on evening out the heat. I believe it would be quite the same in a one pipe system with all the vents closed on the off cycle. Bear in mind that with traditional 2 way rad vents the rads condensing the fastest simply pull air faster from the room on the off cycle - the opposite of what you really want them to do.

Obviously controlling this central venting manifold would require a control on the PLC level and is a steam enthusiast kind of project. But the cost of that would be very small in relation to the cost and trouble to install the vent plumbing. And the knowledge required to program the simple PLC really isn't much. If I had a one pipe system I don't think I could keep myself from attempting this - I believe the end result would be better than a 2 pipe system. I would do the control before adding the pump. I still haven't added a pump to my system because just the natural vacuum was so effective - and totally free with no maintenance.

I thought the reason for a pump was to achieve steam temperatures in the 140-160F range on demand?

It is. I'm merely pointing out that the plumbing required to install a pump opens up very significant venting control possibilities that I couldn't resist taking advantage of. And to achieve your 140-160 degrees on an old system of any size with all its leaks will likely take quite a pump. Maybe its just me but I wouldn't have my heart set on a pump from square one. Work with the natural vacuum first and get a feel for it before investing in a pump. Just a suggestion.

ChrisJ

PMJ said:

ChrisJ said:

PMJ said:

Just a few thoughts for those with one pipe systems considering this. While significant vacuum induced mechanically with a pump makes good sense, simply the achievement of routing all the rad vents with some sort of tubing to a central location would open the door to all kinds of possibilities even without a pump. System balancing now done with different fixed vent rates at each rad would now become dramatically more flexible by simply opening and closing each vent tube for different lengths of time relative to each other at a central manifold. Also possible would be turning off the heat completely in various rooms on some daily schedule. And if all vents close when the burner goes off there is significant natural vacuum to take advantage of. We can all appreciate that very tiny differences in pressures within a system cause the steam to flow in different directions. I have observed in my two pipe in natural vacuum that when the burner is off the vacuum forms slightly faster in rads where the steam is condensing faster and pulls more from the mains on the off cycle than the others. I turns out that this is exactly what I want and has a significant impact on evening out the heat. I believe it would be quite the same in a one pipe system with all the vents closed on the off cycle. Bear in mind that with traditional 2 way rad vents the rads condensing the fastest simply pull air faster from the room on the off cycle - the opposite of what you really want them to do.

Obviously controlling this central venting manifold would require a control on the PLC level and is a steam enthusiast kind of project. But the cost of that would be very small in relation to the cost and trouble to install the vent plumbing. And the knowledge required to program the simple PLC really isn't much. If I had a one pipe system I don't think I could keep myself from attempting this - I believe the end result would be better than a 2 pipe system. I would do the control before adding the pump. I still haven't added a pump to my system because just the natural vacuum was so effective - and totally free with no maintenance.

I thought the reason for a pump was to achieve steam temperatures in the 140-160F range on demand?

It is. I'm merely pointing out that the plumbing required to install a pump opens up very significant venting control possibilities that I couldn't resist taking advantage of. And to achieve your 140-160 degrees on an old system of any size with all its leaks will likely take quite a pump. Maybe its just me but I wouldn't have my heart set on a pump from square one. Work with the natural vacuum first and get a feel for it before investing in a pump. Just a suggestion.

I think a vacuum pump will always be out of the question for me.

However, I keep thinking of a setup where every radiator has a return going right back to the inlet side of the radiator isolated by a trap. The inlet has a valve to control the feed of steam and on the end of the radiators are vents with checkvalves.

ChrisJ

What do you think?
Single pipe vacuum system controlled by 12 or 24v thermostats in each room?

PMJ

I'm not picturing this - could you make a sketch? Do you mean you want to drain the rad from a different place and route it back to the one supply line you have?

Pumpguy

@swei, The key to choosing a vacuum pump is it's suitability for use on a "wet" application. This means it's ability to handle water vapor and possibly slugs of water carryover. The usual water jet vacuum producer and liquid ring vacuum pumps used on steam heating applications do this very nicely.

I understand Igor has used the type of vacuum pump you suggest, but I cannot comment on it's long term suitability. I believe he has reported more than 2 years with no problem.

The air capacity sizing standard I like to use is 1 ACFM / 1K EDR. Others in the industry frequently use 0.5 ACFM / 1K EDR. Ultimately it depends on system air tightness and temperature of incoming air/vapor mixture.

@Chrisj, As I understand it, the original intent of the Paul system was to have a way to have mechanical air removal on a one pipe system. With mechanical air removal, the system doesn't "breathe". There is always a vacuum present on the air outlet of each Paul vent on the radiators. When the Paul valves cool and open, there is a vacuum present for any air to exhaust into. The boiler doesn't have to push the air out with each steam cycle.

I don't think there was much consideration for the additional benefits of vacuum like being able to make steam at a lower temperature, or improved heat transfer qualities. At least there is nothing I have read to suggest there might be.

Please keep in mind that I'm a pump guy, and by no means any kind of expert with boilers, system controls, and that sort of thing.

Pumpguy

@pjm, see attached file. A Paul air line system is a conventional one pipe system, except there are separate air lines on the outlet side of the radiator vents. These air lines are piped down to a vacuum pump, and air is mechanically removed, rather than using steam pressure to push the air out with each steam cycle.

jumper

I like ChrisJ's idea but he'll have to tilt his rads the other way?
If anybody is serious about making any one pipe into pure vapor heating,then they should try my idea first because it's the easiest.Plug all vents and attach a fitting anywhere that it's convenient to evacuate air.Then evacuate as much air as you can when boiler is off.

If there is no air then there can't be any air binding so deep vacuum should not prevent condensate return. If some room is too warm simply cover part of the radiator.

PMJ

Pumpguy said:

@pjm, see attached file. A Paul air line system is a conventional one pipe system, except there are separate air lines on the outlet side of the radiator vents. These air lines are piped down to a vacuum pump, and air is mechanically removed, rather than using steam pressure to push the air out with each steam cycle.

Thanks for this @Pumpguy . I see how it works.

My thought is to have individual lines from each rad all the way to a tank/manifold at the pump to be able to control the venting of each rad individually.

PMJ

ChrisJ said:

What do you think?
Single pipe vacuum system controlled by 12 or 24v thermostats in each room?

I'm not quite sure how that would go - would the steam travel to the discharge side of the trap too and cause problems? Maybe not. I'm really not sure. Maybe other folks will comment. I need to consider it a bit. Interesting idea.

Pumpguy

As technology progresses, there is always the opportunity to improve on an earlier concept. I'll stay with what I know best, pumps, and leave the control aspect to others.

Just today, I visited a high school first built in the early 1900's and added onto several times through the mid '60s or thereabouts. The resident engineer that manages the heating system was telling me during mild weather he has trouble heating the gym. After looking at the system and talking things over, we came to the conclusion he just needs more steam in the gym's fan coils than the rest of the school. Now, how that can be accomplished is a problem for others to address.

SWEI

Pumpguy said:

@swei, The key to choosing a vacuum pump is it's suitability for use on a "wet" application. This means it's ability to handle water vapor and possibly slugs of water carryover. The usual water jet vacuum producer and liquid ring vacuum pumps used on steam heating applications do this very nicely.

I'm curious how "wet" the air in a Paul system is when compared with the typical "air over water" situation in a condensate tank?

The air capacity sizing standard I like to use is 1 ACFM / 1K EDR. Others in the industry frequently use 0.5 ACFM / 1K EDR.

Given the cyclic nature of evacuation in a Paul system, would an accumulator of some sort make any sense? I'm imagining a sort of reverse pressure bladder tank that could be "charged" using a small pump running longer cycles.

SWEI

BTW, how deep a vacuum does a Paul system actually require in order to evacuate the air?

The controls guy is getting intrigued by the 'solenoids on manifold' concept...

ChrisJ

SWEI said:

BTW, how deep a vacuum does a Paul system actually require in order to evacuate the air?

The controls guy is getting intrigued by the 'solenoids on manifold' concept...

Wouldn't the answer be "how low of a temperature do you want to run?"

What do you want? 212F? 200F? 180F? 140F?

SWEI

That depends on whether (and how well) the system can hold vacuum. As others pointed out above, we're mixing our metaphors between a vacuum/vapor system and a Paul air removal system.

ChrisJ

SWEI said:

That depends on whether (and how well) the system can hold vacuum. As others pointed out above, we're mixing our metaphors between a vacuum/vapor system and a Paul air removal system.

If you're just trying to remove the air I'd bet even a slight vacuum would likely work assuming you get a slight negative pressure to all radiators.

PMJ

ChrisJ said:

SWEI said:

That depends on whether (and how well) the system can hold vacuum. As others pointed out above, we're mixing our metaphors between a vacuum/vapor system and a Paul air removal system.

If you're just trying to remove the air I'd bet even a slight vacuum would likely work assuming you get a slight negative pressure to all radiators.

Exactly Chris. But I will add that the single most significant thing vacuum brings to the party is simply not letting air back in on the off cycle. And you don't need a pump for that. I think what many miss is that on the conventional system the moment the burner goes off, air from the rooms fills the vacuum created in each rad by the collapsing steam. All forward motion of steam from the boiler through the mains to the rads stops instantly and the rads totally fill with air. When you simply prevent this the vacuum created in the rads continues to pull steam from the mains and boiler with the burner off. In my 2 pipe I have a Magnehelic set up between the header and the dry return. It shows 1-2 inches water lower pressure in the dry return than the header for the entire time the burner is off (10-15 minutes depending on conditions) of my 20 minute 3cph cycles. The forward motion of the steam never stops (even if it is slow) and the mains have no air. I mind you this is in conditions when I actually need heat. In very mild conditions I do finally go off long enough to fill completely back up with air from leaks.

It turns out that when you simply stop having to push out all that air every time the burner comes on things get remarkably easier. The truth is that what is really needed is not so much fast moving steam but continuously moving steam - like what was the case when these systems had coal fired boilers. Venting becomes a non-issue. If it is in the 20's outside after my first preheat run in the morning of about 20 minutes recovering from setback my average burn time after that varies between 5-8 minutes evenly spaced in 20 minutes cycles (3CPH). The first call for heat won't be satisfied till about noon depending on the sun. Steam is always in forward motion, venting a non-issue. I have one vent on a 1000EDR system.

I would say that what the Paul system does is eliminate the need to remove air once the system is hot and running. If the system is tight it is not removing much each cycle - there isn't much to remove.

Crank it down, spread it out.

SWEI

According to this, the Paul System actually relied on both.

BTW, it looks like @izhadano may have gotten started by looking at the Paul System http://forum.heatinghelp.com/discussion/132043/Paul-system

Pumpguy

@SWEI, your question about system vacuum is the very question I asked when I started this thread.

Also, many thanks for your post on Paul System from the early 1900's. This is very interesting and much appreciated.

I know with absolute certainty the Monodnock Building in Chicago is still heating with a one pipe Paul system. Theirs has 4 separate branches with an individual vacuum pump serving each branch. Over the past 25 years or so, I have sold them several liquid ring vacuum pumps, and converted them from being sealed with once through city water to now using recirculated condensate.

In my experience, a weak point of any vacuum system, regardless of how the vacuum is achieved, is the check valve(s) between the vacuum side, and atmosphere. These don't last forever. As they begin to leak, a vacuum pump will begin to cycle more often, while still maintaining system vacuum.

Hopefully eventually someone will notice the vacuum pump is short cycling and come to the conclusion the air check valve is leaking and needs to be repaired or replaced.

jumper

Pumpguy said:

In my experience, a weak point of any vacuum system, regardless of how the vacuum is achieved, is the check valve(s) between the vacuum side, and atmosphere. These don't last forever. As they begin to leak, a vacuum pump will begin to cycle more often, while still maintaining system vacuum.

Yes which is why one needs a wet seal or positively operated valves. Plural intentional.

SWEI

Pumpguy said:

The air capacity sizing standard I like to use is 1 ACFM / 1K EDR. Others in the industry frequently use 0.5 ACFM / 1K EDR. Ultimately it depends on system air tightness and temperature of incoming air/vapor mixture.

Are these sizings for Paul systems?

What kind of vacuum level range does the pump operate at on a Paul system?


thank you~

Pumpguy

@SWEI, These sizing standards are used for 2 pipe vacuum return systems. I would think these would be appropriate for a Paul system too, although I have not been able to find any sizing standards for one pipe Paul air line systems.

In my experience with vacuum pumps on steam heating systems, frequently the vacuum pump chosen is too small for the load. That's why it is difficult to err in providing too much air capacity, and 1 ACFM / 1K EDR works for low vacuum systems. Two ACFM / 1K EDR is recommended for variable vacuum, sub-atmospheric systems.

if the vacuum pump is running start-stop, controlled by a vacuum switch, the size of the pump, it's air removal capacity in terms of ACFM, would determine the running time to cycle between the settings of the switch. The air tightness of the system and temperature will also affect the vacuum that can be achieved.

Again, the question of operating vacuum range on a Paul air line system is the question I posed when I started this thread. At this point, no certain and specific answer has been posted.

I would be willing to bet that if this question were posed to most suppliers of vacuum pumps for steam heating applications, they would not be able to provide an answer from experience.

ChicagoCooperator

Pumpguy said:

I know with absolute certainty the Monodnock Building in Chicago is still heating with a one pipe Paul system. Theirs has 4 separate branches with an individual vacuum pump serving each branch.

Probably off topic, but I'm dying to know, if you know whether the Fine Arts Building on Michigan still uses their system that way?

Pumpguy

I am not aware of any details of the Fine Arts Building's heating system.

With business activity real slow right now, I just might try to find out. If I learn anything, I'll post it here.

ChicagoCooperator

Thanks, I have a few radiator pictures, but can't find them right now. Nope, found em!

ChicagoCooperator

Here they are....

Pumpguy

Hmmm... Certainly looks like a Paul air line system to me.

Tried calling the bldg. office to see if I could get contact info for bldg engr, but no answer today. I'll try again Monday.

ChicagoCooperator

I've seen that system somewhere else and have pictures of it, but can't find them or remember where.... Let us know if you find out any further information.

Pumpguy

So far, building office confirms they're still on steam heating. Haven't been successful in contacting bldg. engineer. I'll continue to try as time allows.