Low pressure downstream of PRV station

Tony Conner_2

... the PRVs have the right springs in the pilots? Can you hear the steam whistling through the PRVs?

Why the two-stage drop? We'd do that in one stage for the pressures you're working with. What's the steam load?

Darrell Lackey

Mystery of the Disappearing Steam Pressure

In a district heating system, central boilers deliver 125 psig steam to multiple buildings through insulated direct bury site piping. All buildings but the closest smallest building operate correctly over a wide range of boiler plant psi settings, and even in the problem building, the first stage pressure reducing station works fine. Each building receives between 70 psi and 100 psi steam, depending on the boiler setpoint. Steam enters each building through a steam separator, parallel first stage PRV's (1/3 and 2/3 with a globe valved bypass taking steam from 100 psi to 60 psi), then through a 2nd parallel PRV stage (1/3 and 2/3 with globe valved bypass again that reduces from 60 psi to 12 psi). Central plant steam replaced existing low pressure (12 psi) boilers in each building, so plant steam connected to the header downstream of old boilers and old boilers were valved out of the system. Steam is used for space heating, no process loads.

The first stage PRV in the closest problem building works fine. Whether steam enters at 70 psi or 100 psi, if it's cold or warm outside, it leaves at 60 psi as designed. The second stage PRV cannot deliver more than 5 psig whether inlet pressure is 60 psi, 75 psi, or the 1st stage is bypassed altogether and inlet pressure is 100 psi. When 2nd stage PRV's are bypassed, there is no problem delivering pressure from the 1st stage station. When the valve directly downstream of 2nd stage PRV's is closed, pressure directly downstream of the PRV's reads 18 psi.

PRV's were checked for capacity and correct operation by the manufacturer. Capacity is similar to the existing building load with customary safety factor. The existing boiler cycled on/off even at moderately low outside air temperatures. 95% of the loads were valved off, one of the two parallel PRV's were removed, then the other, and still max. pressure downstream was 3-5 psi. That removed the possibility that 1st stage valves were undersized and starving the 2nd stage, that the building was clamoring for too much steam, or that the 2nd stage PRV's were undersized. Temperature of condensate entering the receiver was 132F so there wasn't bypass occuring. Existing pipes were checked and no evidence of lost steam was found. It was only when 2nd stage PRV's were set to reduce from 60 psi down to 58 psi that downstream pressure raised to raise slightly above 5 psi. Two pressure gauges were installed downstream of the 2nd PRV station, one directly downstream and the other about 20 ft. Both have the same readings.

The mystery is where did all the steam pressure go.....

Brad White_9

Interesting

Darrell, the one constant factor in this situation is the PRV's if I gather correctly.

Bypassing them via the globe valve gives you pressure, either one alone or apart does not meet setpoint (12 psig I assume). So what is left are the second-stage PRV's themselves, my gut tells me that because "the solution is around them."

That closing the outlet gives you 18 psig over such a small volume of pipe is maybe a little odd but conceivable. Boyle's law and all that; when that surplus of pressure is applied to the full LPS header volume, pressure drops to 5 psig or so. I take it that these are pilot operated valves such as Spence ED's? And that the isolated section downstream includes the pilot sensor port? I could see that over-shooting a bit in such close quarters.

The one thought I have is that the valve may be the correct size but may have reduced seats or the pilot settings may be off. Either or both conditions.

Global question: Setpoints and pressure requirements aside, how does the building heat? And what does the steam serve? If CI radiators you can do with a lot less than 5 psig. If VIFB coils or other air handling unit coils with control valves you probably want at least 5 psig ahead of the control valve as is often recommended.

My point on this is, if the building is heating correctly, maybe things are just meant to be?

Darrell Lackey

Reply to T. Conner

We'll have the manufacturer double check the springs. They weren't comfortable taking steam from 100 psi down to 12 psi in one step. Steam load is about 3,100 lbs./hr.

Thanks.

Walt Deacon

Diaphragms

Also assuming Spence ED's, remember the pilot is the PRV, and it delivers pressure to the bottom of the main valve which is really a big diaphragm operated "globe valve".
When you isolated the downstream load, the pilot could have passed enough steam to control the pressure. It does this through the tubing that vents downstream. So it sounds like your pilots are working.
The main valve is actuated by wafer thin diaphragms in the base of the valve. The pilot squirts steam pressure to this chamber through the other piece of tubing.
BOTH of those pieces of tubing are supposed to have restrictor ports in them. This keeps the pilot from slamming the main valve, and keeps the DIAPHRAGM FROM FAILING.
Two things can happen regarding these restrictor ports. They can plug, or go "mysteriously" missing. It sounds like they may be missing either because they plugged, or maybe never installed. The tubing looks like standard tubing, so if the little cardboard envelope was missing at installation, an enterprising individual may have used standard compression fittings. The valve would work for while, then do exactly what you describe... and because of those cardboard envelopes, this happens a lot!

Brad White_36

I agree with 2-stage

Many times you can do it in one stage (Lord knows I have and from higher initial pressures too), but two stage is quieter and gives better control at the second stage. At 3100 PPH the valves are probably fairly small first stage- 1"/1.25" I am gathering for 1/3-2/3.

Another general question, are the pilots set the same (wrong springs or not)? Do they ride up and down in parallel?

Or do you stage them based on outlet pressure by using different pressure settings for each pilot? Just curious.

Tony Conner_2

If...

... you're happy with the results, then you're all set. I routinely drop 125 down to 10 in one stage, and use the smallest PRV that will handle the peak.

Safety valve sizing is always an issue, but once you have series & parallel PRVs are in the picture, they can become HUGE (and expensive) and tricky to size. A lot of folks seriously undersize their safety valves after PRV stations because they are looking at the flow under normal operating conditions, not worst-case. If the safety valve settings up and downstream of the PRV are used (and they should be), the differential pressure across the PRV is typically considerably higher than the operating conditions. As a result the possible flow can be far more than the flow required just to make things work correctly.

Tony Conner_2

If...

... you're happy with the results, then you're all set. I routinely drop 125 down to 10 in one stage, and use the smallest PRV that will handle the peak.

Safety valve sizing is always an issue, but once you have series & parallel PRVs are in the picture, they can become HUGE (and expensive) and tricky to size. A lot of folks seriously undersize their safety valves after PRV stations because they are looking at the flow under normal operating conditions, not worst-case. If the safety valve settings up and downstream of the PRV are used (and they should be), the differential pressure across the PRV is typically considerably higher than the operating conditions. As a result the possible flow can be far more than the flow required for operational reasons.

I'm not suggesting that noise and redundancy shouldn't be a consideration when sizing PRV stations, but many people don't understand the domino effect it has on safety valves.

Brad White_9

Excellent point, Tony

When upstream valves let go (rarely thankfully) the through-put would overwhelm a net-size safety valve.

Given a choice of cost, I would use one-stage as you suggested if noise and finite control were not an issue. Good to have your perspectives!

Thanks
Brad

Darrell Lackey

Reply to Walt

Thanks for your input.

Darrell Lackey

Reply to Brad White

Thanks for your ideas. We're having the manufacturer verify the setup.