28 replies to this topic

[processji]

Request your experience on steam condensate system that is in design phase.

 

We have total 25,000 kg/h of condensate at 3.99 bar g and 151 deg c. The condensate is from 2 re-boilers. Flow from each re-boiler is 12500 kg/h. Each re-boiler has condensate 4" line combining into 10" line.The 10" is routed 1600 meter to the de-aerator  operating at 0.3 bar g. The de-aerator is in a location which is 50 m below the location where the re-boilers are located. The 10" pipe is schedule 40.The De-aerator inlet nozzle is at 23 m from grade.

 

The line is routed via pipe rack which has elevation of 10 m. There are bends and turns etc.The 2 phase in the 10" line from re-boilers to de-aerator will lead to hammering and erosion in the line.To me this system is not a good engineering practice. Rather the condensate can be flashed in a drum which can be located close to re-boiler and condensate pumped to the de-aerator. This will enable disengagement of steam in the drum and flash steam tied into inlet steam line to re-boilers.The condensate can be pumped off, thereby avoiding the ill effects that comes with long 2 phase line.Also pumping the condensate will enable to do away with 10" line and a smaller line will do.

 

My question is ,is my understanding right.

Edited by processji, 26 November 2013 - 03:21 PM.

[Bobby Strain]

Yes, you are correct to pump the condensate with such long piping.

 

Bobby

[curious_cat]

 

Rather the condensate can be flashed in a drum which can be located close to re-boiler and condensate pumped to the de-aerator. This will enable disengagement of steam in the drum and flash steam tied into inlet steam line to re-boilers.

 

The flash steam will be at a lower P than reboiler inlet , right? Can it be tied into the inlet? I don't think so but I may be wrong. 

[DB Shah]

The flashed steam will be at lower pressure hence cannot be floated with steam control valve up stream, however can be connected at control valve down stream, configuration is attached of the system.

Attached Files

•  Reboiler condensate.xlsx   13.35KB   121 downloads

[katmar]

If the steam to the reboiler is controlled by a valve on the inlet (as shown in DB Shah's sketch) then there must be a steam trap on the condensate outlet from the reboiler.  The pressure drop through the steam trap means that the flash steam is at a lower pressure than the pressure in the steam chamber of the reboiler.  It is therefore impossible to simply route the flash steam to the downstream side of the control valve.  However, the flash steam is often reused by placing a thermocompressor downstream of the control valve and drawing the flash steam into the suction side of the thermocompressor.  Of course this can only work if your steam supply pressure is high enough.

 

The alternative method of controlling the steam to a reboiler is to put the control valve in the condensate outlet and having the full steam pressure on the steam side (with a varying condensate level and therefore varying surface area in the reboiler).  But this also results in the flash steam pressure being below the supply steam pressure and the same argument as above applies.

[DB Shah]

Dear Katmar,

The idea is not to recycle the flashed steam but to balance the drum and re-boiler chest pressure. The condensate in the drum is all the time saturated liquid. The condensate drum has no vapour outlet, it is just a hold up volume with balancing line.

 

There is no steam trap, condensate line is dipped in the drum as indicated in the sketch.

 

This system is in operation in one of our plant and working fine since last many years.

[katmar]

Dear DB, the impression I got from the OP and earlier posters was that the aim was to recover and reuse the flash steam and I thought you were proposing the same thing. Sorry, I know your postings well enough to realise that you would not make a mistake like that! I suppose in effect the condensate drum and pump in your system function as a high capacity steam trap.  Are you able to run two reboilers off one condensate drum like this, or does each reboiler have its own drum?

 

But this system still leaves the OP with the problem of the condensate flashing as it is transported the long distance to the deaerator. Somehow the condensate must be cooled before it is pumped and flashing it to generate extra steam is a good way to do it.

[DB Shah]

Dear Katmar,

We have only one drum for both the re-boilers. There is no steam trap in the system.

 

When I saw this configuration for first time, even I was under the impression that parallel operation of both the reboilers may be difficult, one chest pressure will interfere with another, particularly when one of them is operating at very low load.

The balancing line is small 1" dia. The beauty is that when "A" reboiler is at full load and "B" is at lower load, the condensate from "A" will enter drum and flash, this steam will enter "B" and balance requirement will be fulfilled by control valve of "A". There will be a dynamic balance of chest pressure of both the reboilers.

 

Transferring the condensate at deaerator is not a problem, the pump will discharge at higher pressure (well above saturation) and the control valve of drum (LIC) will be just near to deaerator so that condensate flashes just in dearator. However as the line length is 1600 mts, lying instrument cables will be costly. In that case a HC/PIC can be provided at dearator inlet to maintain single phase at upstream.

 

In our case, we have LIC near drum and deaerator is just 300 mts away.

[curious_cat]

When I saw this configuration for first time, even I was under the impression that parallel operation of both the reboilers may be difficult, one chest pressure will interfere with another, particularly when one of them is operating at very low load.

The balancing line is small 1" dia. The beauty is that when "A" reboiler is at full load and "B" is at lower load, the condensate from "A" will enter drum and flash, this steam will enter "B" and balance requirement will be fulfilled by control valve of "A". There will be a dynamic balance of chest pressure of both the reboilers..

 

The two control valves, what are the loops set to control? 

 

I'm still confused as to how this works: Take your example, if B is to be run at low load and the inlet valve on B throttles to try and reduce steam pressure in reboiler B. 

 

But if A is introducing full pressure steam won't this pressure be communicated via the balancing lines to reboiler B as well? 

 

Thereby circumventing the efforts of B's controller to throttle back on its steam pressure? 

 

What gives?

[curious_cat]

Also, what's the advantage in doing it this way rather than just two steam traps & two inlet valves throttling steam?

 

PS. To clarify, I'm not questioning your strategy just trying to understand it. 

Edited by curious_cat, 27 November 2013 - 06:52 AM.

[processji]

Dear all

Thanks for the response. Putting back flash steam to steam line is not in current design. My error in stating so.

Flash steam recovery is not the driver in our project. The de-aerator is 1600m away and we need to transfer the condensate to the de-aerator. As highlighted by Mr.Katmar ,the control valve is on the outlet condensate line. The flash steam as one option can be condensed and put back to the condensate vessel and pumped to de-aerator.

In the current design the condenste from reboiler outlet is piped to de-aerator. The line will be 2 phase and we are suggesting to add vessel near to reboiler and flash the condensate.

 

I feel 1600m 2 phase line is not a good idea.

Thanks

Regards

[curious_cat]

Does your current solution involve a condensate pump or is it a gravity line? I wasn't clear on that. 

 

If it does have a pump can you add back pressure enough to prevent 2 phase flow? Or is a pressurized condensate return not an option. 

 

Just wondering. Not sure if this is realistic or not. 

[processji]

The case is real.

As stated in the initial thread to this topic.

Current design does not have condensate vessel. Outlet of re-boiler there is control valve. The line is directed to re boiler 1600m away.

as stated by you a adding a vessel and pump would avoid 2 phase.

I am seeking views on whether it should be the way forward.

[DB Shah]

Dear Katmar, I am working out a schematic which may convey the system in a simpler way. Will come back soon.

 

@ curious cat, condensate inlet nozzle on deaerator is at @ 24 mt from grade. Normal operating pressure of the steam is @ 3 Barg, considering DP across CV 0.5 the available head for condensate return will be @ 25 mt, deduct DP of line and trap to this and we will not be able to push the condensate into the deaerator and hence the system.

 

@processji,

If your concern is not steam conservation then what you have pointed out in your post#11 will work. In one of our plant I have designed similar system. The condensate is returned to polishing unit ~ 1.5 km far. The drum vapour is condensed in a condenser, uncondensed vapour is vented to atmosphere without any control or isolation valve. The only caution you should take is DO NOT PUT ANY ISOLATION ON the condenser or condenser vent. This is to avoid any vacuum generation in the drum and ensures that all the time drum is floating with atmosphere.

 

However this will be a costly alternative. There will be capital cost of drum, condenser, cooling water piping, level control valves of drum and instrumentation. Also you will incur operating cost of condensate pumping power, cooling water circulation apart from energy loss of the steam

 

If present header is posing problems like erosion, increasing header size can be a cheaper alternative.

 

Attached Files

•  Reboiler condensate-atm vent.pdf   35.9KB   88 downloads

[katmar]

Dear DB, I look forward to seeing your schematic.  I have not had time to think it through completely, but I think there is a good chance to be able to do the transfer without any pump - even with a back pressure valve at the end of the line to suppress flashing.  Note that the condensate is at 4 barg and the deaerator is at only 0.3 barg.  The 25 m3/h in a 10" line gives negligible friction if there is no 2-phase flow.  There is also a 50 m drop in elevation which will help, but I am a bit confused about whether the 23 m elevation of the deaerator inlet nozzle eats into that 50 m or not.

 

@processji - to help DB Shah to create a relevant design can you provide a sketch with the actual elevations of the major items.  Also, I suspect it will be pertinent to his design whether there are any high or low points along the 1600 m line.  Please describe this or include it in your sketch.

Edited by katmar, 28 November 2013 - 04:45 AM.

[DB Shah]

@Katmar,

I am attaching a file with 3 sheets showing following cases-

 

1. Normal case ie design case

 

2. Lower load of B reboiler.

In this case as the chest pressure reduces, small amount of vap will flash from drum and enter B from balancing line. This flow will be governed by amount of flash from condensate of A as well as limitation of balancing line for differential pressure of 0.3-0.25=0.05 Bar. If flashed flow is 100 kg/hr through balancing line and required steam to B is 3000 kg/hr, The control valve will provide 2900 kg/hr. As the chest pressure of "B" is low there will be a condensate level build up equivalent to 0.05 Bar ie @ 0.5 mtr in condensate line.

 

3. Imagine a case of Very low load of reboier "B" (20% turn down, normally turn down will be 40~50 % of column)

 

In stand alone condition when the reboiler is not connected to drum, the chest pressure required may be 0.05 Barg. (assume)

Here little more condensate will flash and as available pressure drop across drum and "B" is higher compared to case-2 (0.3-0.05 = 0.25 bar). Assume the flashed rate to be about 500 kg/hr. In this case balancing line limitation may govern the flashed steam to "B".

 

As the difference between "B" chest pressure and drum is 0.25 Bar, condensate build up will be ~2.5 mtrs. In this case the tubes may get submerged in the condensate and hence cut the surface area. This will now call for higher chest pressure because as area is cut off, LMTD needs to be improved.

 

In next step chest pressure will build up to 0.1 bar, the rate of flashed steam from drum to "B" may reduce to @400 kg/hr. the liquid level will reduce, allowing little more surface area.

 

Thus a dynamic equilibrium is achieved with flashed flow of 400 kg/hr and balance from control valve.

 

This seems to be complicating in explaining, I hope I was able to convey it to you.

 

 

 

Attached Files

•  Reboiler condensate.xlsx   33.52KB   55 downloads

[processji]

Dear Mr.katmar and Mr.Shah

At the outset thanks a lot for sharing your thoughts and time.

Shall sketch and send it.

I am travelling and do it soon.

regards

[DB Shah]

Yes with available differential pressure, pump is really not required. Proper sizing of header can solve the problem. As this is 2 phase exact ups/downs data will be required. Schematic will be helpful.

 

@processji

If you have an isolation valve near deaerator on this condensate line, you can throttle the valve and observe the hammering on the line. If this helps the you can install either a HC valve or install additional globe valve at next available opportunity.

[processji]

 Condensate system.docx   13.81KB   28 downloadsDear  Mr.Katmar/Mr.D.B.Shah.

Please find attached sketch. Red line is 10" portion sch 40.De-aerator operates at 0.3 bar g.

regards

[katmar]

Processji, thanks for the flowsheet.  This confirms what DB Shah wrote in post #8 and what I wrote in post #15.  If you control the pressure at the junction of sections 1, 2 and 3 to be just above the saturation pressure corresponding to the temperature there using a control valve located at the inlet to the deaerator, then as long as there are no high points along the way the pressure will be above the saturation pressure all along the line.  The friction losses in such a large line with liquid only are affectively zero, and the static pressure will increase as the line drops down to the deaerator site. 

 

I think this is the ideal solution.  No pump is required, and because the steam is not flashed at source all the energy is returned to the deaerator.  The only downside is the potentially long control cable - but if you are using a bus system there may be cables in place already.

[DB Shah]

Dear Processji,

 

Based on your iso I have sketched the pressure profile and modification in attached sheet.

You can install one PIC just near to deaertaor which will control upstream pressure of condensate header at @ 4.5~ 5 Barg ie above saturation pressure. Also with this additional pressure drop, existing reboiler outlet control valve DP needs to be reduced. The pressure profile will be some what as below-

 

At outlet of reboiler 4.0 Barg saturated condensate at 50 mt elevation

Control DP 1 bar

Pressure at control valve outlet 4-1 = 3 Barg, (there will be flashing and hence 2 phase flow in 4" header)

As condensate flows down to 10 mt elevation, it will gain head of 50-10 = 40 mt approx 4 bar gain. (2 phase will become single phase as head gained)

Pressure at pipe rake of 10 mt = 3+4-frictional loss(assume0.2 bar) = 6.8 barg

Assumed frictional loss of 0.5 bar on pipe rake 10" header upto deaerator location.

Pressure at 10 mt elev pipe rake at deaerator location =6.8-0.5 = 6.3 barg

Condensate flows up to deaerator nozzle at 23 mt. head loss = 23 -10 = 13 mt

Pressure at new PIC upstream = 6.3-1.3 ~ 5.0 Barg

Pressure of deaerator = 0.3 Barg

Specify pressure drop in PIC = 5-0.3=4.7 Barg - some margin (there will be flashing in d/s specify in CV datasheet)

 

You can see that 2 phase occurs only after flashing in reboiler CV & d/s of new PIC.

1600 mt 10" header will be single phase.

 

Hope this helps.

Attached Files

•  Isometric condensate.xlsx   18.84KB   49 downloads

Edited by DB Shah, 30 November 2013 - 06:29 AM.

[Bobby Strain]

All this discussion assumes that all the equipment and piping is existing. If it is a design yet to be installed, you should let us know.

 

Bobby

[katmar]

Thanks to Bobby Strain for the reminder.  The very first line of this thread says "...system in the design phase".  It is a bit like the riddle that starts off "You are driving a bus..." and ends with "What is the bus driver's name?"

 

I agree with the proposal put forward by DB Shah in Post #21.  The only change I would make would be to decrease the line size from 10" to maybe 6" for cost saving.  25 t/h in a 6" line gives a velocity of only 0.4 m/s and a pressure drop of 0.15 bar over 1600 m for single phase flow.

 

It is important that the controller at the end of the line is not sensitive to changes in flowrate, and for this reason I would not make the pipe smaller that 6" even though there is more than enough pressure drop available.  The controlled pressure must be made up only of the discharge pressure from the reboilers and the static head.  It should be influenced as little as possible by the flow rate and therefore the friction losses.  This will keep the 2-phase region at the start of the line almost the same length, regardless of flow rate.

 

The saving in switching from 10" to 6" would probably be more than the cost of the extra controller, and would eliminate the original concern of hammering and erosion.  Avoiding 2-phase flow is always a good principle in my book.

Edited by katmar, 01 December 2013 - 06:47 AM.

[curious_cat]

 The only downside is the potentially long control cable - but if you are using a bus system there may be cables in place already.

 

@katmar

 

Which is the long control cable? In the sketch @DBShah attached the PIC at the deaerator inlet has a sensor local to it, right? 

 

Or did you have a different control strategy in mind?

[curious_cat]

@DB Shah:

@Katmar:

 

I had an idea. Wasn't sure if it is practical or not. Here goes:

 

The reboiler outlet condensate valves will in  the current scheme lead to flashing in the 4" pipes. 

 

What if those control valves were moved further down those 4" pipes? If a elevation drop of ~10 m is gained before the valves are encountered we solve the flashing problem entirely right? No more flashing even in the initial pipes.

 

Or is this scheme unrealistic on operational grounds? 

 

What's your opinion?