2 replies to this topic

[dtmgo1]

Hi, I am working on basic engineering for a pressure relief system, we have decided to install high and low pressure flare headers for high and low pressure relief systems, with high and low pressure flares. The high pressure flare header will relief vessels placed in 300 ANSI class service, and the low pressure flare header will service the tanks. How to determine the pressure threshold to send miscellaneous relief loads to one or another header? I suppose that this would depend on the substances in question, the backpressure that may exist at each point in the header (constant or superimposed backpressure). Botht the low and high pressure flare headers are supposed to be for emergency use only.

A second question is how to calculate the flow of liquid into a flare knockout drum (flare knockout scrubber). The vessels are those found in oil batteries (separators, free water knockout tanks, treaters, etc), so to size them for blocked flow, fire, etc, I think of vapour only exiting the PSV, but how can i estimate the maximum liquid flowrate that may enter the flare knockout drum? I can use a process simulator to see how much gas will condense due to cooling in the header, but this seems to be not a conservative approach. I have seen some people using the inlet liquid flow into the vessel, but this seems excessive and unrealistic, but if high level switches-alarms fail, then the liquid will travel downt he flare header.

Any thoughts appreciated.

Regards,

[ankur2061]

dtmgo1,

Some guidelines regarding liquid carryover to flare KOD based on a reputed international company are provided below:

The objectives of a flare knockout drum are:
1. to separate liquid from the gas before it is disposed;
2. to hold the maximum amount of liquid, which can be relieved during an emergency situation.

With the above it is important to recognise that the maximum gas relief case need not coincide with the maximum liquid relief case. This means that the size of the knockout drum shall be determined by both the maximum gas relief case as well as the relief case at which a maximum amount of liquid is relieved.

Also it is important to note that If the relief flow is an "absolutely dry gas", the possibility could be considered of relieving this gas through a separate (absolutely dry) gas relief line connected directly to the flare, bypassing the knockout drum.

Two more categories are considered for flow to the flare KOD according to the company design practices

The second category is recognized as "Essentially Dry" which means that there is a very small amount of entrained liquid as droplets in the flow to the flare and the KOD design should consider this.

The third category is recognized as "Two Phase Flow" where significant quantities of liquid are present in the flow to the flare system. This is possible when pressure relieving and depressurizing devices are releasing hydrocarbon gas which contains significant quantities of heavier hydrocarbons (C4+) which can drop out as liquids when the pressure is reduced across the relieving or depressurizing device. In such a case the Flare KOD design will have to take into account this "2-phase flow" to the KOD.

Now you need to define what type of a discharge to the flare system you are expecting from your system and decide accordingly the design of the Flare KOD.

Hope this helps.

Regards,
Ankur.

Edited by ankur2061, 24 April 2012 - 10:43 AM.

[Robert Montoya]

dtmgo1:

The primary design problem with a low pressure system is in delivering the off-gases to the flare. In many cases the pressure drop involved in moving the gas to a safe area for the flare, combined with the pressure drop across the flashback prevention devices, can easily exceed the total pressure available to the system. In other cases the pressure available may, with care, be sufficient to move the off-gases to the flare.

A low pressure system may be sufficient to feed the flare, provided the header lines are sized adequately to reduce pressure drop. It maybe necessary to perform the pressure drop calculations manually as many of the standard computer programs for line pressure drops will not work at such low pressure differentials. Low pressure systems should be checked to be sure that back pressure imposed by the flare header does not interfere with proper operation
of relief devices.

When system pressure alone is inadequate, a common solution is to add an off-gas blower to the system after the off-gas source, but prior to any safety device. This allows the system to develop adequate pressure without the need to modify the sources, reduces the size of the flare header and permits the use of the flare safety device of choice. The disadvantage of this method is its inherent safety problem combined with system reliability. The blower is handling flammable or combustible material. Care must be taken in choosing a blower and motor that will minimize the possibility of off-gas ignition.

Reliability can be improved by having a blower spare installed and by performing routine preventive maintenance as required. A backup catbon adsorption system may also be used where only short periods of blower downtime are anticipated. The system pressure drop should also be checked to make sure that the back pressure during high release conditions does not exceed the unit's pressure rating. The alternative to adding a blower is to design and operate the system at a sufficient pressure to feed the unit off-gases to the main flare header^ or, the low pressure stack can be run separately from the main stack and supported by the same structure.

Another design consideration for low pressure flares is the possibility of extinguishment by high velocity winds. Special flare designs maybe required and are available for certain installations (e.g., offshore oil platforms).

Headers—Sometimes it is economically warranted to separate the high and low pressure flare systems. Generally, the lower pressure reliefs dictate the size of the relieving system. In some cases, however, the controlling quantity comes from relief devices set at high pressures and in these cases the high pressure system, with a smaller piping network, might be sized separately.