Steam Electric Plant, Condenser Vacuum Loss Rate on Trip

The rate is pressure increase over unit of time.

That, in more precise terms, is what I'm looking to see if anyone here has knowledge or experience with real system responses.

Not looking for precision, just something like, for instance, would half the operating vacuum be lost in minutes, hours, days?

I've heard that some condensers have installed vacuum breakers to hurry the process along. But if one is not in a hurry, say for restart purposes, these wouldn't be used?

Wouldn't that depend on the size of the turbine, which hasn't been stated, and a range of local factors?

There are too many unknowns to give you a useful range of rates of vacuum loss when gland seal steam is lost.....

If the venturi/ejector can eliminate non-condensables as fast as they are leaking in, then vacuum can be maintained.

If non-condensables are leaking in slightly faster than the rate of removal, and the system is large relative to the imbalance, the vacuum can be kept for a very long time.

If the leak rate is high, and steam to the venturi/ejector has also been lost, then vacuum will be lost rapidly.

Each system is a little different, so you won't really know... until you do.

Basi

I'd assume that the steam jet air ejectors are lost as well, so it's just a passive turbine-condenser system with ~8 shaft labyrinth seals without sealing steam (and thus passively air in-leaking) and a turbine-condenser system free space of ~150,000 cubic feet.

So I guess it's a question of expected shaft seal in-leakage rates under these conditions.

The labyrinth seals should leak 3-6 scfm..best guess

Laby seal leakage can be easily calculated if you know ambient conditions. Somewhere in my dusty archives I have several papers on calculations methods (I used to do this stuff) but my guess if you Google it you can find it.

That being said, without more detail we can't really say if leakdown will take hours or days. The answer has to be: It depends (on more than what we know so far).

Plants also have mechanical vacuum pumps in addition to steam jet air ejectors. If it is desired to maintain vacuum, the mechanical vacuum pumps do it as long as they have electrical power and there are no major failures causing abnormal air in leakage.

The decrease in vacuum will follow an exponential curve tending to atmospheric pressure in about five time constants, with the time constant determined by the size of the condenser and the equivalent area for air to leak in.

In most cases if the trouble can not be immediately resolved and the turbine restarted then one would engage the jacking gear to rotate the turbine while cooling down. One would continue to operate the ejectors and provide steam to the gland sealing system to maintain vacuum while the turbines cool down in order to prevent rotor warpage and damaging your rotor blades.

Your basic hypothesis appears to be more of a mathematical issue then an engineering issue

If the boiler/turbine control operator is a very good operator, he would be opening up the vacuum breaker valve to AVOID steam seal in-leakage! When you have steam seal in-leakage, there is a problem with warping and/or cracking of the labyrinth seals which can cause big problems in the near future. It is highly suggested that you open the vacuum breaker valve ASAP to diminish the vacuum in the turbine when you lose your seal steam supply to avoid these problems.

If you have an emergency shut down of your turbine your main concern is for your rotor not your sealing labyrinths. Besides, you always have live steam being injected into your labyrinths from your sealing steam system. Your LP turbine exhaust end is always under a vacuum and must be provided with a source of sealing steam. Generally it is supplied by the leakage from your HP glands but even during start up live steam is fed in from the steam system to the labyrinths otherwise it would be impossible to pull a vacuum in the condenser.

I'm afraid that your comment lacks experience in this matter.

Quoted from original post:

When a large steam power plant is tripped and isolated from the turbine condenser system, (and thus gland sealing steam is lost), at what rate is condenser vacuum lost due to seal inleakage?

Quoted from Spinco:

Besides, you always have live steam being injected into your labyrinths from your sealing steam system.

My response:

Due to the original post having been written as such, the loss of gland sealing steam was mentioned and steam stop valves etc are shut as the unit was tripped therefore the rotor is already decreasing in speed is implied. I agree with your assessment that the priority is the rotor. Perhaps you failed to read the part about the loss of sealing steam (which implies that even your leakoff has diminished as you have a tripped turbine on your hands).

As to the quote from Spinco, obviously as per the OP, that is not the case.

I have seen times when a turbine has been tripped and there was a loss of sealing steam. The turbine had tripped and a loss of sealing steam had occurred. You will obviously try to maintain lube oil pressure, etc. In the case of lost seal steam as the original post mentioned, it is a very good idea to open the vacuum breaker to avoid the air in-leakage to protect your labyrinth seals. By opening the vacuum breaker, you will also reduce the speed of your turbine more quickly.

Quote from OP in Reply #2:

That, in more precise terms, is what I'm looking to see if anyone here has knowledge or experience with real system responses.

Not looking for precision, just something like, for instance, would half the operating vacuum be lost in minutes, hours, days?

I've heard that some condensers have installed vacuum breakers to hurry the process along. But if one is not in a hurry, say for restart purposes, these wouldn't be used?

My Response:

It is because of the loss of seal steam supply that I highly recommend the use of the vacuum breaker for the purpose of avoiding the cooler air coming across the labyrinth seals as air in-leakage. If I wanted a restart, I would by all means diminish the vacuum in the condenser and turbine. I would rather take the time to replenish the vacuum as opposed to the lengthy outage required to eventually replace the labyrinth seals.

As this #2 Reply was taken into consideration for my earlier post, I feel that I should hereby reiterate so that the OP can understand more perfectly what I was trying to say to him.

You may also note that the Vacuum Breaker Valve control is interlocked and should not operate until your turbine is below a set speed. Generally by this time any steam for seals has diminished from your leakoff, etc. In answer to the original poster, once your vacuum has begun to rapidly diminish, you are obviously having issues with air in-leakage. Hopefully your hand is on the vacuum breaker control switch trying to open the vacuum breaker valve. I hope this answers your concerns.

Ron

Frankly the whole question that was purposed seems a theoretical problem posed by a student for some sort of paper as opposed to a valid concern by an operator. Basically as you point out he is stating that "a large steam power plant is tripped and isolated from the turbine condenser system". Where did all that steam go? It disappeared in a flash?Since the gland sealing system is always backed up by system steam it should take quite a while for all that steam to be lost since the main user is the turbine and that supposedly was also tripped.

I would still rather go for slow leakage through the labyrinths rather than a huge rush of cold air before I could engage the jacking gear and get a slow cool down of the rotor.

Based on the OP's reply in #2, I felt that he was not a student, but a trainee or operator looking for an answer. I gave my reply based on standard operating instructions. As I mentioned earlier, you will not be able to open the vacuum breaker valve at first as there are interlocks based on turbine speed which should make it impossible to do until the time is right (except in the case of miswires). Your turbine should slow to around 1200 RPM or less (estimate) before the interlocks will allow the vacuum breaker valve to open. Until this time, your turbine should have enough steam to be self-sealing. Below this point, your vacuum will deteriorate at an increased rate indicating that you already have a lot of air in-leakage. It is advised in the operating instructions to open the vacuum breaker to open the vacuum breaker ASAP in an effort to avoid catastrophic damage to the labyrinth seals.

As to your concern about the increased airflow from opening the vacuum breaker, the vacuum breaker valve admits air into the condenser about 20 feet below the LP turbine blading making your concern a non-issue. As the OP's comment in #2 suggests, he is looking for real system responses, I chose to answer as I did with emphasis on safety of the equipment especially the labyrinth seals which are expensive to replace and require a lengthy outage to accomplish change-out.

I have seen some cracking of labyrinth seals in the past so I know that it happens. I have also seen turbine blading damage, but it was obvious that it was damage due to turbine water induction and not as a result of evacuating the vacuum through the vacuum breaker valve.

I'm the original poster of this question, but had a heart attack on Monday so wasn't able to check in.

I'm home now, and besides surviving, finding all of this discussion here was some of my best news.

I'll go though the discussion, pass out some GA's and other feedback.

In any case, thanks a lot.

N&P

To all:

I'm using a reply to myself, but I think this gets distributed to all contributors.

This is a question from a somewhat theoretical safety analysis, and regulatory perspective (think nuclear).

• Steam is lost from the reactors because of isolation valve closure.
• Little or no operator action would be credited (except maybe if it could make things worse). Electric power loss may be assumed as well.
• In this case, in-leakage to the turbine condenser for a time would be better than out-leakage therefrom.

• Discussion in #6 [ballpark leakage/seal?]; #7 [calc approach]; and #8 [exponential behavior] suggest that evaluation may be useful and practical.

The RG2 - Spinco back and forth discussion was also very interesting on typical operator actions and the potential operator use of vacuum breakers, for this and other scenarios

N&P

N&P... You wrote:

'...In this case, in-leakage to the turbine condenser for a time would be better than out-leakage therefrom....'

.

Is this related to the '...(think nuclear)...' insertion a few lines before? If not, what if your reasoning?

.

If you were basing the preference for in-leakage over out-leakage on release of radioactive material, there are a couple things to consider:

1. In the typical PWR, if there is contamination in the steam plant, there is a much bigger problem to contend with... namely a primary to secondary leak. In that case you could have contamination all over the plant and leaking out at steam traps and seals.

2. 'Out-leakage therefrom' is not an alternative to 'in-leakage'. 'Out-leakage' cannot be an alternative to 'in-leakage' because, starting with some vacuum in the condenser, leakage out of the condenser cannot begin to occur until sufficient amount has leaked in so that there is not a greater pressure outside than inside the condenser. So choosing 'out-leakage' over 'in-leakage' is not really an option.

3. Once vacuum has been completely lost in the condenser, 'out-leakage' still lacks a the requisite difference in pressure to occur.

How you don't take offense to my nitpicking.

BWR in this case, but the same idea.

If leakage past MSIVs is to a volume that is at negative pressure, then leakage out (to the environment) cannot occur until equilization.