Webster Steam system with vacuum pumps

Double D

Pictures are difficult due to the crammed space this pump is located in but I was able to get a picture. 

jumper

>>That said, this is a common error I see. When just vented to atmosphere, that part of the returns would not see any vacuum.<<

System versus component assembly? How about storing degassed water in vented tank?

Double D

In addition to the multiple issues with this system, they have a control specialist installing Belimo zone valves on the run-outs leading to the convectors. The Belimos are replacing the old pneumatic valves. When you look at the pictures, remember I said on the run-outs leading to the convectors.

Double D

I finally did get a chance to see the system run. The maintenance man I originally dealt with was not on duty. The boiler pressure is well over the 5psig they thought it was running at. All of the return lines in the building ranged from 168deg to 187deg. The vacuum pumps were off. The only good news is the 52cfm vacuum pump does have 5hp motors.

EBEBRATT-Ed

@Double D



If you can't find any info on the AHU capacity there are several ways to approximate this. Buffalo Forge is still around they should be able to tell you.

or

Measure the coil face area, count the number of tubes and return bends and the tubing size and the coil thickness. If they are not return bend coils, They could be straight tube non freeze steam coils the same applies without return bends.

Then you need the fan capacity in CFM which a balancer could measure for you.

Then you need the temperature of the mixed air entering the coil this depends on the building design and the % outside air and return air.

That's a long way around.

You could also disconnect the condensate line and catch the condensate in a barell and time it to calculate the lbs/hour

Double D

@EBEBRATT-Ed Thank you. Those were a couple of ideas I wasn't thinking of. Outdoor air mix seems to be typical on all Buffalo Forge blower coils. Percentage differs from broomstick to collapsed folding chair height. The newer style York blower  coils have motorized dampers. 

Double D

This blower unit had a new coil installed. They said when it was installed, it hammered a lot. They came back and repiped it and the hammering stopped. I wonder what the configuration was before this one.

Double D

@Pumpguy

  At times it's difficult tracing pipes since so many of them are buried behind walls. In the pics there are 2- 2" pipes grouped into one. It appears the one leads to the suction side of the primary vacuum pump which is actually the feed tank inlet in the boiler room. Does this sound correct?

EBEBRATT-Ed

The two pipes teed together with the valves and check valves in the first two pictures are pumping out of the condensate pump going somewhere. You can tell from the way the check valves are installed.

The bottom picture looks like a vacuum return as it has a compound gauge on it. That looks like return water to the condensate pump

Double D

@EBEBRATT-Ed
That return line you see entering the vacuum pump is under the basement floor between pits. It appears that underground line would be full of water. A dry return from one side of the banquet hall which doesn't heat is connected to that pipe below the floor. You can see the connection of the 1" pipe at the pump inlet. If that underground pipe is full of water, how can that section vent? 

DanHolohan

If the return is underground, you have to size the vacuum pump to lift that water when it starts, along with the vacuum you need for the rest of the system. The air goes along with the water.

Double D

@DanHolohan

If the air moves with the water then there must be something preventing the pump from doing both. The banquet room has several radiators on one side that haven't heated for years. It appeared the dry return wasn't venting so I cracked a union on the dry return and the radiators heated for the first time since the 90's. Now on to figure out why it won't vent.

DanHolohan

The vacuum pump is supposed to start with the burner. If it's sized properly, it will lift the water from the pipes that are below its inlet and the air will follow. That requires more power than a vacuum pump serving a system without a lift. Could it be that the vacuum pump wasn't sized with enough capacity to do this job?

Double D

That's definitely a possibility. I would like to first get past the excessive operating pressure. Then take care of the numerous failed traps in the building. I have multiple issues. The vacuum pumps are just one of the many problems. The good news is the pressure is now down to 2psi from 7psi. They thought the building was going to stop heating when I dropped the pressure.

Pumpguy

If I see the lift pipe correctly, there may, or may not be a lift fitting incorporated in the tee, just downstream of the ball valve, as in figure 3 in the attached file.

If this is a constructed lift fitting, at least when the vacuum pump is running this low return line will get completely drained of condensate.

If not, the best that can be obtained is most of this low return will remain filled with condensate with just a minimum air space at the top of the horizontal run.

Either way, the only way this low return can be drained, completely or partially, of condensate is for the vacuum pump to run continuous. When the vacuum pump shuts off, the pressure becomes equalized and the condensate just lays in the low return line.

The proper way to pipe a low return arrangement like this is to use an auxiliary accumulator tank with float switch as shown in the attached file.

When the vacuum pump is operating, the auxiliary accumulator tank just acts as an oversize lift fitting and condensate just flows right through it.

When the vacuum pump is not running, condensate flows by gravity into the auxiliary accumulator tank until enough condensate accumulates to lift the ball float tripping the switch causing the vacuum pump(s) to run start - stop to empty out this accumulator tank, ensuring the upstream return lines are drained of condensate.

Double D

 There is another pit in the basement floor where that pipe below the floor takes its drop down. Definitely enough room for an auxiliary accumulator tank. If that pipe below the floor is filled to the top with water, there's more system not seeing vacuum. From the looks of it that secondary pump in the photos takes care of all radiation in the banquet room. The first pump i posted in the pit is also handling radiation in the banquet room. 

EBEBRATT-Ed

@Double D

Really need to see the system as a whole to know what is going on. Vacuum systems are complex

Pumpguy

@Double D, For the other pit in the basement, you could use a pump as a mechanical lift. These are probably a better choice if the lift is more than 5 feet or so, or the available downstream vacuum is not enough to do the job.

The attached file shows the correct piping arrangement.

Specifically, this lift pump's receiver tank should be vented to a vacuum return line and not just to atmosphere.

When vented to a vacuum return line, these low returns become vacuum returns too, not just atmospheric returns.

An auxiliary accumulator tank is only used for the low returns that flow by gravity to the lowest point in the system, and just upstream of the vacuum condensate return pump.

Just keep in mind to move condensate, you need a pressure differential. This can be gravity, a mechanical pump, or a difference in pressure where downstream pressure is lower than the upstream pressure.

Double D

@Pumpguy

 I will do some more checking but it's possible that pump was moved when an outdoor elevator was moved to an indoor elevator in that area. If it's true that pump was moved, there's a chance it isn't powerful enough to do the job. The pits are 35' away from each other. When I ask how long has the banquet room been a problem, all they can say is a long time. Now I'll ask them if the problem started around the time of the elevator work. As always thank you for all of your input. 

Pumpguy

@Double D, Not sure what you mean by "it isn't powerful enough to do the job."

Typical pump discharge pressure rating on units like these is 20 PSI. This should be posted on the data plate on the panel cabinet. I wasn't able to magnify the picture to see what's on the nameplates.

I don't think plus or minus 35 feet is going to affect the condensate flow that much. If it does, it will only mean that it takes a bit longer to cycle between the on and off settings on the float switch.

The real issue with this installation is how the low returns are piped as I commented on earlier.

Double D

For now I only have this blurred picture of the rating plate but it does look like 20psi. The fittings in the pictures are the only ones that look like lift type. (nothing like you attached file) The vent on this tank is also atmospheric.

What you mention in an earlier comment is what appears to be happening. I assume this set up has never worked.

\/ \/
[ If this is a constructed lift fitting, at least when the vacuum pump is running this low return line will get completely drained of condensate.

If not, the best that can be obtained is most of this low return will remain filled with condensate with just a minimum air space at the top of the horizontal run.

Either way, the only way this low return can be drained, completely or partially, of condensate is for the vacuum pump to run continuous. When the vacuum pump shuts off, the pressure becomes equalized and the condensate just lays in the low return line. ]


I do see a vacuum reading in the gauge when the main vacuum pump it running and this one is completely off. I will need to be there to observe the cycles but from the description I'm getting from the maintenance men is this pump is only cycling to discharge its condensate. The pump just basically short cycles.

Pumpguy

@Double D Check the settings on the selector switches. They may be set to the FLOAT ONLY position. If so set, the pump will only operate start-stop controlled by the float switch.

With the existing return line piping arrangement, the vacuum pump needs to operate continuously to provide the needed pressure difference (suction) to lift and remove condensate.

When the vacuum pump is off, there is no pressure difference (suction) so the condensate just accumulates in the low returns.

This accumulated condensate forms a water seal which prevents the system from venting, so steam can't get into the radiation and they stop heating.

Double D

Thanks again. That would explain why the radiators heated when I cracked the union on the dry return. I will do like you say and run the vacuum pump in a continuous mode to see what happens. 

Double D

@Pumpguy

Running the pump continuous seems to have worked however I had to shut it down because it was taking on steam from failed traps. There were 2 blower coils that were still not heating. They were filled to the top with water. In the end an auxiliary pump may be the answer. I will give a progress report after I get past the failed traps. 

The main vacuum pump in the boiler room seems to run very short cycles and stays off for long periods of time while the boiler is running. 

Although the system still has hammering, they mentioned yesterday that it's a Lot quieter than it used to be.

Thank you so much for your input thus far. It has been extremely helpful.

Pumpguy

The main vacuum pump in the boiler room seems to run very short cycles and stays off for long periods of time while the boiler is running.

What are the settings of the vacuum switches on this unit? Do they sense the vacuum on the receiving tank, or from some other location?

If the vacuum pumps shut off, either the vacuum switches need adjusting or something else is controlling them.

These vacuum switches can be tricky to adjust. Its done on an adjust and try basis. The attached file explains the process.

For a vacuum source, I like to use a small hand squeeze type vacuum pump like an auto mechanic uses to bleed hydraulic brakes.

Double D

At the present time the only gauge in the boiler room used to read vacuum is located on the basket strainer at the inlet of the boiler feed tank. The Square D vacuum switches on the vacuum pump are definitely controlling the pumps. Since there are no gauges on the vacuum pumps lead and lag pumps, I was going to install them to get an idea of what vacuum those switches are seeing. I suspect possibly the basket strainer may have a build-up. Nobody can remember anyone ever checking. Although I have not taken an accurate reading of the vacuum produced by the system, there is definitely vacuum present at the basket strainer when the pump is off. For some reason it's not reaching through to turn the vacuum pump on.

Pumpguy

Even though this system doesn't have the auxiliary accumulator tank which, IMO, would solve a good many of the described problems, there are a few changes that can be made to improve this system's performance.

1) Just to the right of the vacuum gauge on the basket strainer is an adjustable spring loaded vacuum relief valve. For this application its in the wrong place.

As with an auxiliary accumulator tank, the vacuum relief valve should be piped to allow air to bleed in to the lowest point in the system. The valve should be adjusted to break @ 10" Hg. vacuum.

The purpose of this valve is to allow atmospheric pressure to enter the low return to push the condensate up the lift pipe and into the receiving tank. This is if for whatever reason the vacuum pumps aren't able to create a low enough pressure (vacuum) in the receiving tank so the upstream return line pressure can, by itself, push the condensate up the lift pipe.

Remember, vacuum does not suck, pressure pushes. The vacuum pump lowers the pressure in the receiving tank and upstream pressure pushes the condensate up the lift pipe.

2) Regarding the vacuum switches, on a system like this I like to use a 3 port ball valve for the sensing lines. The common port should go to the vacuum switches while one switchable port goes to the vacuum pump's receiver tank like it does now, and the other switchable port goes to the low return line. This can be piped to the same pipe as the vacuum relief valve. Use large enough diameter copper tubing to avoid short cycling. I like 3/8" OD tubing for this application.

For normal operation the vacuum switches should be sensing vacuum on the low return line, not the vacuum pump's receiver tank. Sensing vacuum off the upper receiving tank can cause the vacuum switches to sense an unstable pressure (vacuum) and cause their operation to be erratic.

Typical settings for the vacuum switches are OFF @ 8" Hg. and ON @ 3" Hg. On a Duplex system like this, there are primary and secondary vacuum switches. The primary or lead vacuum switch would be set at 8 and 3 while the secondary or lag switch would be set at 8 and 2.

The 3 port ball valve allows switching where the vacuum switches sense vacuum from. This is useful when adjusting the switch settings and the main receiver tank is valved off from the rest of the system.

You can pipe the vacuum relief valve and the low return vacuum switch sensing line to the copper pipe that comes out of the return line just upstream of the ball valve. Remove that elbow and install a tee. Off the tee install the vacuum switch sensing line and vacuum relief valve.

Your pictures show a copper tube going behind the basket strainer and down into the trench where the low return line is. Where it comes from and where it goes to is not clear. There may have some attempt to use this tube as a vacuum switch sensing line, or it may serve some other purpose. Just can't tell if its related in any way or not.

DanHolohan

I know of no one in this business who knows more about this topic than Dennis. It’s a joy having him available to answer these questions. Thanks, Dennis!

Pumpguy

Thanks Dan, I appreciate the kind words.

There seems to be some question about how the vacuum pump in this thread functions so here's a basic explanation.

This unit has 2 vertical centrifugal pumps, each of which has 2 separate functions.

Function 1 is to pump condensate to the boiler, boiler feed tank, or wherever, through the vertical discharge pipes.

Function 2 is to pump hurling water through the venturi nozzles to produce the vacuum.

To switch between the 2 functions there are special control valves. These can be operated by hydraulic pilot valves, float switch and solenoid valves, or even direct operated by float and rod linkage.

The receiver tank has 2 compartments, upper and lower. Condensate from the return lines enters the lower compartment from which the venturi action creates a vacuum to lift the condensate up into the hurling chamber.

Condensate recirculates through the hurling chamber and venturi jet producing a vacuum. When the hurling chamber fills with condensate, the upper float switch trips the solenoid valve to open the control valve and send the condensate on to the boiler or boiler feed unit.

Different manufacturers will have variations on this theme but so long as the vacuum producer is a water venturi jet type, the above basic principles will apply.

jumper

>>Different manufacturers will have variations on this theme but so long as the vacuum producer is a water venturi jet type, the above basic principles will apply. <<

Reviewing old Domestic Pump manual those variations seem significant. For larger systems condensate, hopefully degassed by vacuum, is pumped from vacuum chamber. The upper chamber is not degassed below atmospheric if that is where "hurling" water comes from.

I also enjoy learning about this archaic equipment from PumpGuy. Does nobody sell a modern controller for steam heating? Every steam system can benefit from intelligent vacuum generation.
At least I think so.

Pumpguy

@jumper "Does nobody sell a modern controller for steam heating?"

At one time Nash did, but that was in the late 1960s. I have some notes in my SHIPCO catalog describing this type of controller, but notes say "contact engineering". Same for MEPCO.

The little that I know about these systems shows they require a steam throttling valve on the main header along with the controls, which turns out to be a very expensive modification. That, combined with ignorance of the end user and their advisors is enough to kill any interest.

The few that I see these days, and I don't visit job sites nearly as often as I used to, are not operating as originally intended and designed.

I assume this is because some of the control hardware or logic controls are no longer operating, and no one knows how its supposed to operate, much less how to fix it.

retiredguy

I once saw a Vari Vac system by Dunham Bush. The one I saw was not working but after reviewing information on it I thought that that system would be a great temperature controller. My thoughts at the time were that most customers with a vacuum return system would never care for or up-date their systems to allow for this type system to function properly and the cost was prohibitive.

Back in January of 2003 a question was asked on this site by VIV about a Vari Vac System he encountered and commented on by @hot_rod and @Pfitter 10.

I believe that G S Dunham at 718-886-4232 still offer an updated Vari Vac steam return system

By the time that I retired in 2007 most of these vacuum return systems were replaced sometimes with an updated steam system but mostly with a completely new hot water system since most heating engineers have very little experience with steam systems. There were only a few of these vacuum return systems in our service area and most of the customers did not want to spend their money on repairs to something they did not understand. On most of the jobs, the main problem with the vac systems not working was with deteriorated piping or non working steam traps and almost nobody wants to waste money on steam trap repairs. Just let the steam vent blow steam what is the harm is what I was usually told.

jumper

>>On most of the jobs, the main problem with the vac systems not working was with deteriorated piping or non working steam traps and almost nobody wants to waste money on steam trap repairs. Just let the steam vent blow steam what is the harm is what I was usually told.<<

Ideally a correctly working vacuum cycle doesn't need traps. Air is eliminated and vapor is condensed. I think that Lost Art says that often original set up was gravity return but over time traps and pumps are added. Then packaged vacuum units were sold to "improve" those "improvements".

Pumpguy

This is what they're saying at Purdue these days. And they're supposed to be BOILERMAKERS.

jumper

Pumpguy said:

This is what they're saying at Purdue these days. And they're supposed to be BOILERMAKERS.

Talk about "lost art". Somebody ought to reprint an old Plant Engineer's Handbook.

EBEBRATT-Ed

You can't get people with standard two pipe systems to keep their traps up to date, never mind vacuum systems

Double D

I have some better pictures of where the copper tube is attached. I don't see how this pump can sense vacuum where that sensor tube is.  

Pumpguy

The vacuum switch sensing line connection circled in red is correct. This location avoids the pressure fluctuation that occurs due to the pulsing in the main receiving tank when lifting alternating slugs of condensate and air. Also, sensing system vacuum from this location avoids the pressure difference caused by the lift and the switches see actual system vacuum.

I like to run these sensing lines using 3/8" OD tubing and arrange for an unhindered gravity flow of any condensing that might occur in these lines back to the lowest point, which in this case is the low return line.

The 360* pigtail in this line, and the mounting of the vacuum switches with their diaphragms vertical are kind of sketchy IMO. Both are water collection points which should be avoided if possible.

Double D

Is it possible those switches are not working? The only time I see this pump cycle is when the water in the gauge glass is about an inch from the top. When the level drops to one inch from the bottom, the pump shuts off. 

Pumpguy

In my post above from November 17, I mention use of a hand held auto mechanic's vacuum pump to check the operation and settings of these vacuum switches. Any other hand or powered small vacuum pump can be used. You only need to go out to 10" Hg. vacuum. I also posted instructions how to adjust these switches.

Using a separate vacuum pump for testing these switches is a whole lot easier than wondering if the vacuum pumps run continuous because of system conditions, poorly performing vacuum pumps, or bad switches. You've removed 2 of the 3 variables.