10 replies to this topic

[Vishal Sali]

Hi all,

 

I have come across some overpressure cases where the volumetric flowrate of overhead vapors at relief conditions leaving a column system or a separator is less than the volumetric flowrate of overhead vapors at normal condition. However the mass flowrate of the overhead vapors at relief is more than the mass flowrate of overhead vapors at normal condition. So if I compare the volumetric flowrate the scenario becomes not applicable but if I look at the mass flowrate the scenario is applicable. what would be the correct approach??

 

The density of vapors at relief conditions is obviously greater than the density at normal condition.  I have observed this mostly in cases of cooling water failure or even abnormal heat input cases.  Please suggest a path to be followed.

 

thanks

[paulhorth]

Vishalpsali

Volume flowrate at normal conditons is irrelevant because at normal conditions you do not have an overpressure case. The function of a relief valve is to prevent overpressure which can only arise when conditions are not normal.

 

When you say

So if I compare the volumetric flowrate the scenario becomes not applicable

you have not told us what scenario you are thinking of. If it is blocked outlet then it doesn't matter what that volume
flowrate would have been with an open outlet. I don't understand your comment.

 

 Relief valves are sized for MASS FLOW at the relieving conditions. The mass flow through the selected valve must be higher than the mass flow for the governing relief load. This mass flow for the relief load is the nett mass flow to the equipment (in minus out) or generated internally by vaporisation of liquid, at the relieving conditions.  Thermal expansions (of vapour or liquid) are obviously exceptions to this simple rule, because they are volume changes at constant mass.

 

Paul

[fallah]

vishalpsali,

 

Whether a possible scenario can be a credible one depends on the maximum pressure that could be developed by such scenario. If it would be greater than the vessel MAWP, then we will face with a "credible scenario". Therefore, just having lower volumetric flowrate than normal one at relieving condition in a part of the system, here overhead line, shouldn't forced us to conclude that the scenario is not credible.

Edited by fallah, 25 February 2013 - 07:45 AM.

[Vishal Sali]

Hi

 

Thanks for the replies Paul and Naser,  I would like to explain this with an example.

 

This is an abnormal heat input case on a column.  The abnormal heat input will lead to generation of excess vapors in the column system. However all the process flow paths out of the system are in normal operation and open.  So in this case after simulating the column in a simulation software I found that the overhead vapors volumetric flow rate (m3/hr) at relief pressure from the column overhead accumulator drum is less than the overhead vapors volumetric flow rate (m3/hr) at normal conditions from the column overhead accumulator drum. However the mass flow rate of the vapors at relief is more than that at normal conditions. So will this lead to accumulation of excess vapors in the column system.  The overhead line of the column overhead accumulator drum is having a control valve. This control valve is assumed to be at its normal position during the event.

[paulhorth]

vishalpsali,

 

To have a relief case you first have to get the column up to relief pressure. If your overhead vapour control valve can pass all the excess vapour without the pressure reaching the relief set pressure, then that's fine. If the valve is controlling the column pressure it will tend to go open if the vapour production increases. The pressure will not rise until the valve has gone fully open, so first you should check the maximum mass flow through this valve at normal pressure and 100% open.

 

Then you can make the conservative assumption that the PCV doesn't change its opening, and work out at what pressure it will pass the increased vapour at normal opening. If this is still below the relief pressure then this is not a relief case. if not, then the relief load on this conservative basis would be the difference between the vapour mass flow and the mass flow through the PCV. However, this is not likely to be the governing case.

 

Note that for your abnormal heat input relief case, presumably based on the reboiler utility control valve fully open, you should estimate the heat input to the column from the reboiler based on the higher bottom temperature at relief pressure. This higher temperature will reduce the heat input from the reboiler (less temperature difference across it) .

 

You can consider a different scenario, such as failure of the pressure controller, or of the PCV, so that the vapour outlet is blocked,  That will become an overpressure relief case. Two other relief cases to consider are the feed valve fully open, and failure of the condenser..

For each of these,  you should estimate the heat input to the column from the reboiler based on the higher bottom temperature at relief pressure, and normal utility flow.
 

 

Paul

[Vishal Sali]

vishalpsali,

 

To have a relief case you first have to get the column up to relief pressure. If your overhead vapour control valve can pass all the excess vapour without the pressure reaching the relief set pressure, then that's fine. If the valve is controlling the column pressure it will tend to go open if the vapour production increases. The pressure will not rise until the valve has gone fully open, so first you should check the maximum mass flow through this valve at normal pressure and 100% open.

 

Then you can make the conservative assumption that the PCV doesn't change its opening, and work out at what pressure it will pass the increased vapour at normal opening. If this is still below the relief pressure then this is not a relief case. if not, then the relief load on this conservative basis would be the difference between the vapour mass flow and the mass flow through the PCV. However, this is not likely to be the governing case.

I agree with this approach and it seems appropriate. Thanks for the same. 

 

I would also like to know what approach can I follow in case the overhead vapors are going to a centrifugal compressor/positive displacement compressor or a Process Header.  For compressors do we have to compare the mass flow of the vapors at relief condition with the rated capacity of the compressor in mass terms??

[paulhorth]

Vishalpsali,

In my opinion you do not need to design the compressor for the vapour at relief conditions, because there should be a process shutdown before the relief pressure is reached. This shutdown would be activated by a high pressure trip on the column.

Please note that the shutdown can make a difference to the relief case described in the earlier post. You did not mention the compressor earlier. If the excess overhead vapour was discharging through the PCV to flare, then it is valid to take credit for this flow continuing during a shutdown and relief. But if it is going to a compressor which will be isolated on a shutdown, then the relief load would have to be the total vapour, not just the excess flow.

 

Paul

[S.AHMAD]

1. Since the system has PCV at the overhead, I strongly believe that the PSV is sized based on the blocked discharge scenario. The failure of this PCV (e.g. stuck in a close position) can be the cause of overpressure.

2. Please take note that, the sizing of PSV is BASED on actual VELOCITY at the orifice. The velocity could be critical or sonic or sub-critical.

3. For determining the orifice area of the PSV, we can use the mass flowrate or volumetric flowrate Nm3/h. If you notice the equation also contains the P, T, Z and MW of the gas/vapor. The purpose of these operating conditions, molecular weight and the compressibility factor is to transform the mass flowrate into volumetric flowrate at the operating conditions. The actual volumetric flowrate divided by the orifice area, gives the velocity. The velocity that gives Mach No=1 (for critical flow) is the maximum flowrate that can passed through the orifice. For sub-critical depends on the pressure differential.

4. For the same velocity through the orifice, the mass flowrate of N2 will be difference to CO2 but their actual volumetric flowrate will be the same.

5. Therefore, for comparing the the PSV capacity in volumetric basis it must be at the same operating conditions that is the operating at relieving conditions is difference to that normal conditions. That is why we normally compare the flowrate in mass basis since mass does not vary with operating conditions.

Edited by S.AHMAD, 28 February 2013 - 11:23 PM.

[Wheristo]

Hi all..

I happened to come across this topic while searching an answer to what I was looking for... basically in my case I have this overpressure scenario on a column, say a Total Power failure. The consequences eventually lead me to perform a simulation to obtain the relief load. Through simulation I was able to get the relieving rate ("A" lb/hr).

Additionally, In this case I am able to take credit of normal overhead flow ("B" lb/hr) that is assumed to continue to flow downstream. This is obtained from the H&MB. 

 

So my question... when taking credit of the normal overhead vapor flow.. can the flow difference be taken as direct subtraction of the mass flows ("A"-"B")? i.e. the lb/hr obtained from performing the overpressure scenario simulation at relief conditions, minus the lb/hr of the normal overhead flow at normal conditions from the H&MB...

(Sorry if this has been explained in threads above, but I haven't got the full grasp of it.) 

 

I have seen (without really understanding the rationale behind it)... people do the subtraction in volumetric terms (by dividing the mass flows of (A) and ( into their respective densities)... The resulting difference in volumetric flow rate was then multiplied back with the density at relief condition to obtain the mass flow rate that is eventually used for sizing?  

 

What would be the correct way to do this? Appreciate your expert advice.

[paulhorth]

Wheristo (edited - wrong name)

 

Is there an ESD valve on the flow to downstream which takes flow "B"? Is it possible that this valve will close as a result of your Total Power Failure (causing a plant ESD)? If it closes then you cannot take credit for the flow "B" being subtracted from the relief flow.

If thre is no ESDV then you should consider what the flow to downstream is, with the column at relieving conditions. Don't use the HMB value.

 

For example if this flow goes through a control valve, the flow could be higher than the HMB flow because there is now higher pressure at this valve - except if the control valve is a flow controller -  then the valve will close in to keep the flow constant. If it goes to a compressor (which has not been shut down), the flow will be higher than HMB because the compressor suction pressure is higher. You can estimate the flow from the compressor curve. That would reduce your relief rate, so be sure that it is correct.

 

Work out what the plant will do with the column at relieving conditions, assuming no operator actions. Then you can stick to mass flows, and don't have to do that ridiculous calculation with the volumes at diffeent pressures, which has no basis in reality. 

 

Paul

Edited by paulhorth, 23 December 2013 - 02:51 PM.

[Wheristo]

Paul-

Thanks for the reply & advice. That clarifies.