22 replies to this topic

[bosun]

I have two heat shell and tube heat exchangers working in Sulphur Recovery Unit. Detailed process layout diagram and heat exchanger data are attached. Amine acid gas preheater, E1, heats the gas using IP steam. Combustion air preheater, E2, heats the air using IP steam as well.

The odd part I am trying to figure out with these two heat exchangers is that IP steam temperature control valves (TC1 and TC2) are 100% open;however, the IP steam inlet flow rates are much lower than the design value for valve opening of 52%. For example, current steam inlet flow rate going to E1 is 2.70 tonnes/hr but according to design spec. sheet of the valve, at 52% opening, the flow rate should be 2.976 tonnes/hr (this is the maximum flow rate). As seen in the diagram, the two valves are controlled by outlet temperatures of amine acid gas and combustion air. The valve will open if the outlet temperatures are not meeting the required temperature.

I trended valve opening and flow rates from since these exchangers started to operate and found out that the two valves (TC1 and TC2) have always been open 100% but the flow rates have been changed. I also trended steam flow rates, amine acid gas flow rate, combustion air flow rate, and outlet temperatures of amine acid gas and combustion air. As the amine acid gas flow rate increases, the steam flow rate increases but the outlet temperature of amine acid gas decreases. The same trend occurs to combustion air. So I suspect that the heat exchangers are undersized that there is not enough heat transfer is happening in both exchangers but I am not sure if it is that case or if there is other areas I should pay attention. Please give any ideas that can shed some light onto me and to this problem. For more information, I took pressure survey on steam inlet lines for both exchangers and the pressure readings were as expected at around 4300kPa.

Currently, these exchangers operate fine but I am working on this to find ways to increase performance of the exchangers and do optimization.

E1 is one tube pass shell and tube heat exchanger. E2 is two tube pass shell and tube heat exchanger.

Any industrial experts out there please help me!! Thank you!

Attached Files

•  heat exchanger data.pdf   123.57KB   818 downloads
•  PFD.pdf   7.06KB   513 downloads

Edited by bosun, 29 August 2012 - 12:13 AM.

[Dacs]

I think you need to mention your ultimate goals are.

Looks like you want to operate your SRU on a higher capacity (is this right?) so you're studying the possibility of using the heat exchangers at a higher flow.

Have you tried running the exchangers under design conditions? What happened? Did it perform as expected?

While the conditions for your steam indicate no superheat, can you confirm if your steam is indeed saturated? I normally expect that you're producing saturated HP steam inside your SRU from your Furnace Waste Heat Boiler, although it's possible that your plant might employ a superheater to superheat the steam.

The odd part I am trying to figure out with these two heat exchangers is that IP steam temperature control valves (TC1 and TC2) are 100% open;however, the IP steam inlet flow rates are much lower than the design value for valve opening of 52%. For example, current steam inlet flow rate going to E1 is 2.70 tonnes/hr but according to design spec. sheet of the valve, at 52% opening, the flow rate should be 2.976 tonnes/hr (this is the maximum flow rate)

Just to be clear, both your CVs are 100% open and yet it passes less stream through it as designed at 52% opening?

Take a look at your HP steam pressure, it's around 43.2 barG. Is this the pressure upstream or downstream your control valve?

Can you tell me the pressure of the condensate pots downstream your heat exchanger? Does it have a PCV on its vapor outlet line?

To add, can you tell me the inlet and outlet pressures of your steam feed/outlet lines (I suppose you have PGs installed in your heat exchanger)?

Kindly answer these first before we can continue further.

Edited by Dacs, 29 August 2012 - 08:24 PM.

[bosun]

Hello Dacs,

To clarify, my goals are to:
1. Find what the problem is with these two exchangers
2. Once the problem is figured out, find ways to increase performance of the exchangers

So, first I need help to figure out what the issue is with the exchangers.

To answer your questions, the two exchangers did run under design conditions during SRU performance testing and the outcomes were not that bad. However, during the testing the control valves were remained 100% opened.

Also, as you mentioned, the IP steam comes from waste heat boiler and it is saturated steam. The pressure is at around 4300 kPa and the temperature is 255 oC

Both of the valves are open 100% and they have been open since the start-up. The flow rates are even less than the design flow rates at 52% opening. According to design specification of TC1 valve, the maximum flow 2976kg/h and this flow is at 52% opening. However, current steam flow rate going into the exchanger is at around 2700 kg/h.

There are no pressure meters in both inlet/outlet of the heat exchangers as well as on the condensate drums, so I took pressure survey and these are readings I got. For E1, the inlet pressure is 4294 kPa and the outlet pressure is 4306 kPa. For E2, the inlet pressure is 4300 kPa and the outlet pressure is 4324 kPa. These readings were taken on the same day and they are acutal pressure, not gauge pressure. To me, it is odd that the outlet pressures are higher than inlet pressures...

I do not see PCV valves on the vapor outlet from P&ID

I have talked to the instrumentation person and he said he honestly thinks the problem lies with exchangers not the valves. But I am confused that why the valves are 100% open.. There are bypass IP steam lines at the inlets and I checked out in the filed that the bypass valves are closed.

Hopefully I have provided enough information. Please help me figuring out with your expertise. Thank you!

[Dacs]

I have a hunch on what causes the problem but to be sure, do you have a PG on upstream and downstream of TV? If so, can you tell me the readout?

% Opening by itself doesn't mean anything because the flow across a control valve not only depends on % opening but on upstream and downstream pressure.

From the pressure that you gave me, you have a 43 barG HP Steam inlet pressure (from the header) and you have 41.93 barG for E1 inlet. It means that you have (more or less) 1 bar differential pressure across your CV. I calculated the Cv under these conditions (2700 kg/hr and 1.07 bar delP) at 21.3.

Kindly check the control valve nameplate since the rated Cv (Cv at 100% open) is there. Compare the value I calculated with the one listed in the nameplate. If it matches, the control valve is doing its job and it only means that:
1. The designer has undersized the valve during its design, or
2. The designer did his/her work right but your MP/LP Steam header (where the vapor goes from your condensate pots) operating pressure is higher than what the designer has anticipated, or
3. You have excessive pressure drop in your HEx (which is unlikely).

Let me know how it works out

There are no pressure meters in both inlet/outlet of the heat exchangers as well as on the condensate drums, so I took pressure survey and these are readings I got. For E1, the inlet pressure is 4294 kPa and the outlet pressure is 4306 kPa. For E2, the inlet pressure is 4300 kPa and the outlet pressure is 4324 kPa. These readings were taken on the same day and they are acutal pressure, not gauge pressure. To me, it is odd that the outlet pressures are higher than inlet pressures..

That's expected because you have a condensate product and it imposes a static head. Since you have a bottom exit, then there exists some liquid inventory in your HEx that will contribute to the static head.

Attached Thumbnails

• 

Edited by Dacs, 31 August 2012 - 08:31 PM.

[Dacs]

Actually, I took a second look at E1 heat exchanger D/S and the design inlet pressure for your steam is at 41 barG. Since you told me that you have 4294 kPa (41.92 barG) inlet pressure, it means that you are around 0.92 bar higher than expected inlet, and your control valves have to open more to allow the same flow because you have a decrease in pressure drop across the control valve of around 0.92 bar.

This might be the reason why your CV is at 100% open.

[Steve Hall]

Dacs,
What do you think about this possible cause: if the heat exchanger is undersized, no amount of excess steam will improve performance because the condensing steam (and therefore the steam trap) will limit the steam flow. The control valve in this case might not be effective at all. The control loop would be trying to get more heat so it would open the valve, to no avail. It might perform just as well with the valve open 52% as at 100% - the same amount of steam would be utilized because the condensate (resulting from heat transfer area/performance in the exchanger) rather than the steam supply would be the limiting factor. This could be tested by manually adjusting the control valve to throttle to 80%, 70%, 60% and so on then looking at the performance trends - if they don't change then the exchanger, or even the condensate trap, could be the culprit.

[Dacs]

To answer your questions, the two exchangers did run under design conditions during SRU performance testing and the outcomes were not that bad. However, during the testing the control valves were remained 100% opened.

This is what he said, although we can't verify that on our end.

If the HEx were undersized, one possibility is that you're right, the control loop will send 100% output to the valves to send more steam and the condensing steam might limit the steam flow, although if you check the pressure profile, hardly any pressure loss exists:

For E1, the inlet pressure is 4294 kPa and the outlet pressure is 4306 kPa. For E2, the inlet pressure is 4300 kPa and the outlet pressure is 4324 kPa.

I have another idea in mind. Since for sure you'd have a bypass line on the control valves, have it full open (to artificially increase the steam flow) and check your acid gas temperature.

If it heats up, it means the control valve is limiting the flow but if it does not, it might be possible (but not 100% sure) that your HEx is undersized.

Personally, I tend to think the HEx is not the culprit because it should have been caught during commissioning if it were.

Edited by Dacs, 02 September 2012 - 11:30 PM.

[bosun]

Hello,

I have few more questions. Is there any more checks that I can do to make sure that if the problem lies with heat exchangers or the valves such as fouling calculation? I would like to see that if there is fouling issue by doing fouling calculation with actual operating condition and compare with design value. In fact, is there a way to calculate the actual fouling value using operating conditions?

Also, I was not able to find Cv values on the nameplate of the valves. But on the design spec. sheet, there are some Cv values at different valve openings. Could I be able use these values to calculate Cv at 100% opening?

One last question, can heat exchanger give duty that is beyond what it is designed for? For example, during 4 days unit performance testing, the unit operated close to design rate. So, during that time, the maximum duty I calculated for E1 and E2 were largerer than the design duty value. I am wondering if this is possible.

If you can kindly answer my questions, I would really be appreciated.

Thank you!

[Dacs]

Also, I was not able to find Cv values on the nameplate of the valves. But on the design spec. sheet, there are some Cv values at different valve openings. Could I be able use these values to calculate Cv at 100% opening?

It's hard to extrapolate on the Rated Cv based on data points at different % opening because we have to exactly know the CV characteristic to reliably predict the Rated Cv.

I have few more questions. Is there any more checks that I can do to make sure that if the problem lies with heat exchangers or the valves such as fouling calculation? I would like to see that if there is fouling issue by doing fouling calculation with actual operating condition and compare with design value. In fact, is there a way to calculate the actual fouling value using operating conditions?

For this service, fouling will largely depend on your Acid Gas composition. NH3 is notoriously known to deposit as ammonium salts in your piping.

Where does your Acid Gas comes from? I suppose this comes mainly from ARU. Sour Water Stripper (SWS) also produces Acid Gas (which is normally rich in NH3). If your NH3 is high on your Acid Gas, it may be possible to have salt formation in your exchanger.

But this will affect your Acid Gas stream and ultimately the duty of your HEx. I don't think it will play a part in the steam line, which is your concern.

Just to keep your bases covered, you can check for the quality of your BFW, that's an indication of the fouling tendency for your steam.

One last question, can heat exchanger give duty that is beyond what it is designed for? For example, during 4 days unit performance testing, the unit operated close to design rate. So, during that time, the maximum duty I calculated for E1 and E2 were largerer than the design duty value. I am wondering if this is possible.

Of course it can. Either increase the temperature (most likely) difference or increase the Heat Transfer Coefficient (harder to achieve).

Heat Exchangers are usually designed with margins on top of basis of Heat Exchanger size so the increase in duty might come from those margin (most probable cause).

[bosun]

Dacs,

Thank you so much for kindly anwering the huge list of questions I had previously. Unfortunately, I need to bother you again asking more questions! I just want to make sure I cover everything and fully understand the system since I need to present this at the end of my internship. ( I want to do a good job!! )

Here are the questions:
Could you explain again what you wrote below. It is not very clear to me why the valve needs to be open 100%...

Actually, I took a second look at E1 heat exchanger D/S and the design inlet pressure for your steam is at 41 barG. Since you told me that you have 4294 kPa (41.92 barG) inlet pressure, it means that you are around 0.92 bar higher than expected inlet, and your control valves have to open more to allow the same flow because you have a decrease in pressure drop across the control valve of around 0.92 bar.

This might be the reason why your CV is at 100% open.

Do valve vendors provide Cv at 100% opening? I wasn't able to find Cv value from the valve spec. sheet..

If you can answer my questions, I would really be appreciated. And if you have suggestions that I should consider, please let me know.

Thank you in advance!

[Dacs]

No problem, I was in your shoes before and one has to start asking and one has to answer back to further develop our knowledge.

I'll try to explain it in more intuitive terms (a bit tedious and you may find this a bit lecture-ish but bear with me):

As you may know, a flow inside a pipe can happen if you have pressure differential in your inlet and outlet piping and the flow will go from higher to lower pressure and generally the higher your pressure differential, the higher your flowrate.

For a given piping configuration (ie fixed size and pipe length) and fixed upstream and downstream pressure, you can totally define your system and be able to solve for the corresponding flowrate.

This is good but what if you want to achieve a flow lower than the flowrate you calculated above?

In this case, you'd use a control valve, to introduce extra pressure losses to bring the flowrate inside your piping down. And this control valve can actually modulate the flow by varying the % opening.

Control valve is sized for a (1) given flowrate (which is usually at maximum flow) and (2) pressure differential across the valve (ie pressure at control valve inlet/outlet)

When the control valve was designed, the designer assumed 43 barG source pressure and 41 barG Heat Exchanger inlet pressure (which in this case is your destination pressure). Just to be clear though, in practice we don't make Heat Exchangers as an endpoint in our design, but for the benefit of my explanation, I'll make it as one.

Using this data, the designer calculated the Cv corresponding to 2.9MT/hr steam and ~2bar delta P (approximate since you also need to consider piping losses because obviously there's an inlet and outlet piping on your control valve).

But according to your inspection, the inlet pressure of your steam to the Heat Exchanger is close to 42 barG. It means that your destination pressure has increased and you basically have a lower pressure differential across your steam manifold.

Which means one thing: all things equal (% opening of control valve as it was), you'd have a lower steam flowrate. But since the control valve can actually adjust its opening (to allow more or less flow), it will open more to meet the target steam flow to your exchanger.

This has limitations: if your pressure differential is too low for the control valve to catch up, the control valve will open at 100% and yet may not be able to achieve your required flowrate (your control valve acts as the bottleneck). And looks like this may be your case.

The designer has sized the valve for 2 bar pressure delta but your operating conditions only gives 1 bar. Someone has to give in, and that's going to be your steam flowrate.

Do valve vendors provide Cv at 100% opening? I wasn't able to find Cv value from the valve spec. sheet..

They should. Valves are being sold with Rated Cv in mind. It should be on the nameplate but if you can't find it, try to look for documentations for that valve. Try to ask your instrument maintenance engineer in your plant, he should have the documentation for it (hopefully)

Hope this helps

Edited by Dacs, 05 September 2012 - 07:51 PM.

[Steve Hall]

The valve may have a long "model number" listed on a tag or label. The model number is usually made up of a series of codes. From the valve manufacturer's literature the meaning of the codes can be determined; one of them might be the Cv.

I hesitate to add to Dacs's excellent descriptions, but let me point out that when the performance tests were conducted during initial testing the exchangers would have been in a clean condition. You would definitely expect performance better than the design (i.e., fouled) due to the clean surfaces.

[Dacs]

That's true, because heat exchangers are designed with maximum fouling in mind. So if you take this factor out, then you'd definitely have better heat transfer capability.

Thanks for adding that info Mr Hall

I wish you luck with your internship Ms Bosun

[bosun]

Thank you for your thourough explanation.

I checked the valve design spec sheet and it says at maximum steam flow (2976kg/h) inlet pressure of the valve is 4136 kPag and outlet pressure of the valve is 4119kPag. However, currently the steam pressure at the inlet of the valve is 4300kPa which is about 4200 kPag. So this indicates that the valve was designed at lower pressure than the current operating condition (4200kPag vs. 4136kPag). Also, currently there is no pressure drop across the valve since it is 100% opened. Thus, much higher pressure of steam enters the heat exchanger, but at the same time steam temperature is higher than design temperature (255 vs.253). And here I am stuck how to go further...

Also, I have talked to an operation engineer to see if we can open bypass line. He said he already tried opening it and it didn't help. So this leads to me with more questions. I thought about that heat exchanger might be undersized. But as I mentioned previously, we had performance testing conducted after 1 year of operating unit. During testing, we did not run at exactly design rate but close to it. The heat duty I calculated was pretty close to design duty value. So, I thought the heat exchanger is fine. However, from what I heard from the operation engineer, I am not too sure what is causing the problem... I also thought about that steam is flowing too fast that it does not have enough residence time to condese and heat transfer to process gas stream. But here again I am struck how I can prove that the steam is indeed flowing fast..

This project might be easy problem but I am really struggling to get grasp of idea how I should approach the problem. So please I need your help and if you can shed some light onto this I would really be appreciated. Thank you in advance!

[Dacs]

My apologies, I'm starting to get confused right now with the values.

What I know so far is (we'll focus on E1):
1. Your steam inlet pressure is 4294 kPa
2. Your condensate outlet pressure is 4306 kPa

Anyway, can you tell me these:
1. Your HP steam header pressure (upstream of control valve)
2. Acid gas inlet and outlet pressures in your heat exchanger.
3. Post the basis of design of the control valves? (flow/inlet and outlet pressures)

I know you've already provided those above, but it's easier if it's in tabular form, so that my eyes won't get confused much

I made the assumption of having a 43 barG HP Steam header pressure, and now I'm not sure if this is even the case.

Apparently you have these issues:
1. Your steam control valves are not performing as expected and
2. Your Heat Exchangers are also not performing as expected.

This may be two independent or intertwined problems.

For now, we'll try to diagnose the control valve issue and we'll go on with the heat exchanger afterwards

[dchem]

Dear All,

The above interaction helped me too to get better clarity on some of the basics of HT/FF. Thanks for the same.

I would like to give new angle to this discussion. Whether is it right to have TCV on the steam inlet? Wouldn't it be better if it is located on the outlet condensate line? Obviously there will be traps at the outlet, so this TCV can be at the downstream of traps. When TCV gets throttled, condensate level builds-up in the exchanger, rendering some of the HT area useless and thus helping to reduce the temperature. I think this is possible in vertical exchangers only where you have to design the exchangers with some additional area to take care of normal level to be maintained in the exchanger.

One more thought - does it make any difference for location of TCV, whether inlet steam is superheated or saturated?

We can continue discussion on the above system as it is existing one, my question would help the ones who are designing new systems.

Regards and thanks in advance.

[Dacs]

The philosophy of whether placing the CV at inlet or outlet does have its pros and cons.

I've seen both before and engineers would have different views on which one is more effective and as you said, the type of exchanger also determines the suitability of control.

Personally, my preference is to control the condensate line since I feel it calls for better control of heat transfer by varying the surface area in the heat exchanger.

One more thought - does it make any difference for location of TCV, whether inlet steam is superheated or saturated?

The usual thought on this issue is that superheated steam needs to be desuperheated (by sensible heat -> lower HTC) before condensation takes place (by latent heat -> higher HTC).

As a designer, I tend to avoid having superheat (by having a desuperheater upstream of steam inlet) in my design because it may (not always the case though) require additional surface area for sensible heat removal.

But ask the operations people and they will tell you differently. I asked one operations engineer in KSA where I was assigned last year and he said that desuperheaters are just another item in their inspection and they've operated one of their reboilers without it and it didn't give them any issues.

Of course you have other reasons why desuperheaters are ideal (such as avoidance of localized hotspots in your heat exchanger that may cause unwanted reaction such as process fluid degradation), but it's really outside of heat transfer.

Hope this helps

[bosun]

Hello Dacs,

Here is the information you have asked for:

1. Pressure at upstream of the valve is 4294 kPa
2. I don't have pressure meter at gas inlet/out. I was not able to get done pressure survey on gas side due to safety concern.
3. I have attached valve design basis.

I have attached revised PFD drawing. I did not include SWAG stream in previous drawing which combines with AAG stream.

I found rated Cv value at 100% opening. For TC1, it is 224. For TC2, it is 394.

I have a question. The setpoints for steam control vavles are higher than the design outlet temperature. For example, for E2 combustion air design outlet temperature is 220oC but the setpoint of the valve is 240. Is it normal to put setpoint higher than design value?

Hopefully I provide enough information. Thank you so much for helping me out.

Attached Files

•  Valve Data.pdf   96.38KB   494 downloads
•  PFD.pdf   93.58KB   611 downloads

[Dacs]

OK, another confusion.

I thought 4294 kPa is at E1 inlet.

Anyway, I need to have the pressure at the TC1 outlet (TC1 inlet is 4294 kPa as you said in your last post). For your benefit, I'm trying to calculate the Cv of the valve and check if the valve is adequately designed for.

I did a quick check on Cv for TC1 and it matches with what you provided.

I have a question. The setpoints for steam control vavles are higher than the design outlet temperature. For example, for E2 combustion air design outlet temperature is 220oC but the setpoint of the valve is 240. Is it normal to put setpoint higher than design value?

Your setpoint temperature should not exceed design temperature.

Although if I can speculate, they might have set it that high to make the valves fully open at 100%, which is not surprising since the plant is having issues already and it looks like it can't reach that high anyway.

Another thing, ask your plant in how they deal with salt formation in acid gas lines to your Furnace.

[bosun]

Dacs,

There is no pressure drop across the valve. Pressure at the upstream of the valve is same as pressure at the down stream of the valve. I think this makes sense since the valve is fully open.

[Dacs]

There's always a pressure drop inside fittings. Pipes do

Having a valve fully open is one thing, not having any pressure loss is another. Actually, even at 100%, it will impart a pressure drop (depending on Rated Cv).

Upon checking your Rated Cv and did some number crunching, you need to have 0.95 kPa pressure drop for 2700 kg/hr flow. 9.5 mbar is quite low! This is lower than what I expect on a typical pressure drop across a heat exchanger.

I'm inclined to think that (assuming your control valve indeed has the correct Cv) at this point, you have to look elsewhere for the problem.

Something out there is keeping your steam flowrate to go higher, focus on that. Next logical thing to do is to check the downstream (vapor line on condensate drum) pressure. Since you experience the problem in both exchangers, this is what they have in common. Maybe its pressure is backing up the steam and keeping it to flow higher.

Good luck

[bosun]

So then is it fair to say that the valve is not the bottlenecking point? I also think that heat exchanger might be undersized. Can you suggest any idea how I can check if the heat exchanger is right sized? Is there a parameter than I can compare with design spec?

Thank you.

[Dacs]

Apart from doing heat and material balance around the exchangers, I can't think of any.