Turbine-Driven Centrifugal Compressor Operation 5

[EmmanuelTop]

I have experienced something that I consider strange in regular centrifugal compressor operation. I would be grateful if someone of forum members could clarify these issues:

Turbine-driven refrigerant compressor (isobutane) in alkylation unit is suffering from frequent surging when operating in automatic mode. Automatic operating mode means the following:
- antisurge flow controller is in automatic mode
- suction pressure controller is in automatic mode (suction vessel PC connected to high pressure steam servo assembly, which regulates the RPM of compressor by manipulating steam flow into the turbine; steam is condensed under vacuum)
- suction temperature is always constant, meaning that composition of the compressed gas is also unchanged

At minimum alkylation unit capacity, compressor operates at 95% of maximum RPM, developing polytropic head 80% of design value (?). Lowering the RPM pushes the machine into surge region and raises the suction pressure, so the operators found that it is better to run the compressor with almost maximum RPM in manual mode, in order to have relatively smooth operation of the plant. This somewhat causes suction pressure to vary with time, but with no significant consequences.
What surprised me the most is the following:

1) With this parameters I described, antisurge FCV is open 52%. Polytropic head is 80% of design value, as I said.
2) Lowering the RPM from 7000 to 6800 RPM does not affect suction and discharge pressure (?), but it causes antisurge valve to open further, up to 56%! Moreover, machine goes into surge cycles.
3) Switching from manual to automatic mode of RPM control (via suction PC), makes incredible changes in compressor operation: relatively smooth operation is turned into surging cycles, so the automatic operation is completely abandoned.

My questions are:

1) If actual gas composition, suction pressure and temperature are as designed, why cannot we achieve design polytropic head? Is it possible that there is so little process gas (compared to spillback stream), that antisurge flow (52% valve open) pushes the compressor so much right off the curve, developing less head? Is it possible that machine is mechanically damaged, causing lower polytropic head at 95% of design RPM?
2) Why antisurge valve continues to open further when RPM is reduced, if suction and discharge pressures are unchanged? Isn't it contradictory, practically impossible? Less RPM should require smaller recycle stream (if being far enough from the surge point) in order to achieve the same head - that is what I (thought) I knew about centrifugal compressors.

Can you please throw some light on this.
Thanks in advance.

 

[EmmanuelTop]

dcasto,

What is the purpose of TIC controller when all liquid inventory from refrigerant accumulator ends up in the compressor suction drum, in both ways? LIC09 on refrigerant accumulator drum is controlling liquid refrigerant flowrate to the suction vessel. From simple heat balance it is easy to see that is not important how much refrigerant flows through LV and how much through TV, when both streams enter compression suction drum together with spillback gas flow. This means that all condensed refrigerant must end in the compressor suction drum. So I cannot grasp your concept about TIC.

I agree with the proposed PFD modifications in terms of achieving maximum condensing duty by removing LIC10 (LV10) and PDIC valve, and placing liquid refrigerant sub-cooler downstream of the fin-fan condensers. This would maximize refrigerant heat removal at compressor discharge and lower the required pressure for total condensation.

Forget about feed to depropanizer at the moment - I am almost 100% sure it will not function for the next few months, maybe even longer. When reading my last two previous posts, what do you think about C3 accumulation and necessity to have PC valve on the refrigerant blowdown drum? I can't see how can we rid of C3 and provide stable downstream pressure for smooth compressor operation without maintaining constant pressure in the last drum (refrigerant blowdown drum, not refrigerant accumulator drum)?

Onwards,

 

[dcasto]

If you go on high recycle through the spillback valve, the hot gas will mix with the cold liquids from lic09, if LIC 9 isn't open, then you'll get hot vapors into the compressor suction and it's density will decrease and the available head being constant, the outlet pressure will drop and surge.

 

[NghiaPP]

A background of alkylation process may help to understand the system behavior.
In alkylation reactor, C3/C4 olefins react with isobutane to make a high octane gasoline. The process is catalysed by a strong acid like sulphuric acid. The process requires low temperature (0-5 deg.C) and high Ic4/olefin ratio (6-10).
The process is characterized as exothermal, need high recycle of iC4 and need high mixing energy.
Ic4 is a reactant and also a good refrigerant. Therefore an opened refrigeration system like the system in this thread is normal used. The refrigerant compressor is designed to remove reaction heat plus heat added from mechanical equipment (circulation pump, mixing impeller) and keep reaction zone at the required temperature. Recycle iC4 comes mainly from the De-iC4. A part of iC4 is separated from the reactor effluent (in compressor suction drum) before it is sent to the De-iC4. The reaction zone is kept cold by: heat exchange with cold effluent (which is flashed at compressor suction pressure, recycle of cold iC4 refrigerant.

Before going to the compressor, some words about the propane content in the process are needed. Propane comes with the feeds. A small amount can also be made by a side reaction in the reactor (craking). All propane shall be removed from the system to avoid accumulation. It can be done by a purge stream from where it is most concentrate, either at the top of the de-iC4 or at the discharge of refrigeration compressor. In the system shown in the sketch, the discharge gas is partially condensed in the fin-fan, the remaining gas will have higher concentration of C3 and is condensed at lower temperature in the CW cooler. The flow rate of purge stream to De-C3 shall be set high enough to reduce C3 content to a required level (set by the process designer). It will depend on how much C3 entered with the feeds. Higher level of C3 will increase pressure in the refrigeration system (both suction and discharge) to keep the same operating temperature. If the C3 amount is low, the purge can be a gas purge to flare where the loss of iC4 is insignificant.

Back to the compressor, as said above the compressor shall remove the heat from a low temperature (heat source) and reject the heat to ambient (heat sink) at higher temperature. The temperature difference between heat source and heat sink will decide the head of the compressor (compression ratio). The amount of heat to be removed will decide the flow rate of refrigerant to be condensed in the fin-fan.
The relation between composition, temperature and pressure in the suction vessel and the refrigerant accumulator is that the pressure is equal to vapour pressure of the given liquid at the actual temperature. The propane content in liquid plays important roll for the vapour pressure, even at the small change. Given that the propane level is controlled in the allowable range. The temperature & pressure of the suction and discharge vessel is almost fixed.
The compressor operating point (head, flow rate) will be the intersection of the heat curve at given RPM and the system curve (pressure-flow rate relation) at the discharge of the compressor. The head at compressor discharge does depend on the down stream pressure and resistance . The system back pressure is the pressure at refrigerant accumulator plus DP of fin-fan. Since the DP is constant by the action of PDIC. The discharge pressure of compressor shall be constant at all flow rates. The system curve is a horizontal line. However, the compressor head curve will change with the rotation speed. The fan law says that compressor head is proportional with the speed or the flow rate through compressor is proportional with the speed given that other factors are unchanged.
The compressor controller loop (suction pressure – steam flow – speed – flow rate) will work in a certain range, which seems quite narrow. This is a feed back system, if flow rate is less than required, heat removal is low, suction temperature will increase and the suction pressure will increase, which opens for more steam. The input for the controller is suction pressure. It should be in theory equivalent to the temperature of the refrigerant in suction vessel, which is a more direct target of the (temperature of reaction).
If the turn-down causes the refrigerant flow to condensed goes below the surge limit, the surge valve shall be opened to reject the exceed flow. The surge flow does not remove any heat from the system as intention of system designer.
Before answer the questions it should be clear that if the propane level is not in control as it seems to be, it might have effect on the control of the whole system.
- Why can not achieve the design polytropic head ? I assumed the head here means discharge pressure. The head (in meter) is depend on gas properties like MW, Cp/Cv, etc. As mention above, the operating point is an intersection of system curve and compressor head curve. If the back pressure is low, the discharge pressure is also low. It should be constant provide that the down stream equipment is in their function range, check if the PDIC is not total opened or total closed and check the pressure of refrig accumulator.
- Why surge control valve open further when RPM is reduced in manual mode? I have no idea. The more opening does not need a indicator for more flow, if the discharge pressure is not constant.
- PCV at De-C3 feed drum: may be a good ide if the propane import is low such as the propane feed pump is not in function the most of time. I does not recommend to change the layout of CW cooler, unless the feed will be free of C3 in future, in this case the propane feed drum and the pump can also be removed. The current layout can be used to increase the concentration of C3 in purge gas. A new pipe is needed to send the condensate from De-C3 feed drum back to the suction drum by for example a flow control.
Hope it may help to give some light on the problem.

 

[EmmanuelTop]

NghiaPP,

I would not agree that compressor head is dependant on gas properties, as you wrote. It is a function of compressor mechanics and inlet volumetric flow. Only discharge pressure depends on gas properties, but not head developed by the compressor.

Still cannot imagine alky unit design (or any other unit) without compressor discharge pressure control. In spite of having relatively clean service (>90% i-C4 in compressor gas), propane accumulation and downstream pressure buildup will, eventually, cause overspeeding and tripping-off the compressor if suction pressure control is on automatic mode. If nothing else, having fin-fans as discharge condensers will result in different heat duties during one single day, not to speak about months or seasons. In other words, if refrigerant blowdown system is not functioning properly as it is the case now (no liquid pumping to De-C3, because of inadequate pump and De-C3 operational problems with blowdown stream), PC valve on blowdown drum looks like the most simple solution, resulting in minor losses of i-C4 into flare system.

I think it is the first thing we must resolve before continuing to work on compressor control problems.

Thanks,

 

[NghiaPP]

Sorry, I made a confusion, I real mean the discharge pressure is dependent on gas properties.
I mean that the discharge pressure in this system is controlled indirectly via the condenser outlet temperature. What happen when both outlet temperature of fin fan and CW cooler goes down, e.g by 5 deg.C ? The pressure at refrigerant accumulator or back pressure will goes down. It causes refrigerant flow to the fin-fan to increase. This again causes temperature of the suction drum to go down, which makes suction pressure to go down. The controller will reduce steam to turbine until a new operating point is reach (lower discharge pressure with approximate the same refrigerant flow to the fin-fan).
I believe that the instability when running automatic mode is caused by the control loop “suction pressure – steam flow – speed – flow rate”, based on the fact that when the speed is on manual, instability is gone. The feed back of the loop is break in this case. The system still works since surge flow is a mean to fix how much heat is removed from the system, even it is not a optimal operating point.The control system might be not well tuned or operate outside the range it is tuned for. It must be very sensitive if the compressor head curve in the relevant range is flat (a small change in pressures cause a large change in flow). In that case, reduce proportional factor of the controller may help (The bias will increase).
Regards
Nghia

 

[sshep]

Hey Emmanuel,

A vent to flare for controlling C3s will work if you don't have depropanizer feed option. A flow control drag off the refrig acumulator will be effective. This seems to be what the controls indicate now except that the draw is condensed. Any chance you can go to fuel gas insteadof flare?

This has been an interesting thread, most especially the controls. It challenges my mind to understand how many independent variables are available across the whole system. Whenever a system with so much is going on is unstable, it can be hard to figure out if the problem is tuning or a fundamental instability (including overspecification). Can you cover again how we are running the LIC/LV10 and the PDIC?

best wishes,
sshep

 

[EmmanuelTop]

Hi sshep,

LIC10 is always on manual (50-55% open), while PDIC is on automatic mode with 0.4bar set point.

In theory, when purge stream flow to De-C3 is increased (through FCV at pump discharge) LIC/LV10 closes for a few % and pushes more refrigerant gas through PDIC. As a result, level in refrigerant blowdown drum increases to compensate increased purge stream to De-C3.
In practice, I cannot see the reason for not opening LIC10 100% when refrigerant purge stream to De-C3 is not functioning.

When I asked day shift supervisor why LIC10 is not always 100% open, he answered that "system always needs positive differential pressure across the condenser in order to maintain refrigerant flow in the right direction" - which I cannot understand and do not agree with (for example, FCC wet gas compressor does not have hot vapor bypass on discharge condensers). It is also interesting that, according to PDIC measurement, PDIC valve is almost always completely closed in order to maintain set differential pressure of 0.4bar - which indicates that ~100% refrigerant gas flow is going through discharge condensers. In such circumstances, keeping LIC10 partially closed does not have sense at all.

http://antwrp.gsfc.nasa.gov/apod/astropix.html