12 replies to this topic

[Arvind Iyer]

Dear Friends,

We are trying to classify PSVs as "turn-around" and "non turn-around" based on whether they can be isolated and handed over to maintenance in running plant or not.

One strong argument stated that PSVs on Cooling water return lines of exchangers are provided
only for prevention of pressure rise due to valve blocking and hence are not at all needed in running plant. So, they said that all PSVs on cooling water return lines of exchangers can be given to maintenance in running plant.

My argument was that it cannot be given unless there are two PSVs or if the exchanger itself
can be bypassed based on the following reasoning.

The two cases I thought could cause pressure rise are:-

1) Sudden blocking in of the the valves thereby causing thermal expansion of the locked up water.

2) Sudden reduction in the cooling water supply pressure/flow causing vaporisation/excessive
cooling water temperatures.

Can you please help me understand what is the correct reasoning?

Sincere thanks.
Arvind Iyer

[ankur2061]

Arvind,

First of all your post should be in the separate forum named "Relief Devices Forum".

Now coming to your query:

Relief valves in heat exchanger CW (cooling water) return circuits are called TERVs (Thermal Expansion Relief Valves) or TSVs (Thermal Safety Valves). These valves are generally small in size (3/4" X 1" is one of the most common sizes) and operate only when there is expansion of trapped liquid due to temperature rise. The temperature rise occurs only when the cold side is blocked in. TERVs are not meant for relieving vapor and are employed only in 100% liquid filled lines. Hence your 2nd point of vaporization does not hold good. The relief quantities are very small, just enough to relieve the pressure build-up due to thermal expansion.

I have never heard or seen dual TERVs on cooling water return circuits in heat exchangers. There is no practical justification for it. In normal applications such as CW return circuits of Heat exchangers nobody even does a calculation for the relief rate. It is just not worth the time and effort. You just provide a 3/4" X 1" relief valve and leave it at that. One of the most difficult tasks in such kind of applications is to estimate the relieving temperature. The relieving temperature determines the coefficient of expansion of the trapped cooling water and this factor is utilized in calcualting the relief rate. Relief rate calculations for TERVs in most cases is not justifiable.

A very small amount of bleed can be continuously maintained from the CW return circuit to prevent the scenario of thermal expansion if and when your CW TERV requires maintenance and hence dual TERVs are not required.

Hope this clarifies the matter.

Regards,
Ankur.

[Arvind Iyer]

Thanks Ankur.

I was not actually thinking af a bypass TSV provision. I was thinking that bypassing the exchanger from service was required for maintenance.

So, you think that TSVs on CWR lines can be handed over to maintenance team any time in the running plant, right?

Arvind

[astro]

Arvind, the circumstances that you're describing are not clear to me, so apologies if the comments below miss the mark.

If you can bypass an exchanger and not impact the process then removing a TSV with the rest of the plant in operation, should be a simple exercise provided that sufficient isolation is available to permit the work to proceed.

If you're trying to determine whether a TSV must be inspected and tested, then API RP 580 Risk Based Inspection and ANSI/API 576 Inspection of Pressure Relieving Devices would be suitable references to consult for guidance. A risk based approach would provide a means of assessing postponing inspection.

[JoeWong]

Arvind,
Your description is not clear to me.

One of this point confuse me is location of PSV. Is the PSV located within the excharger loop (return line) or the common return line ?

I have the "feeling" (safety can not base on "feeling"), they are located within the excharger loop and there is isolation valve downstream of this PSV.
Base on this understanding, the PSV can be used for the following overpressure protection :
i) Tube rupture scenario where process fluid is on high pressure side and CW in low pressure side
ii) Block-in-heat-on scenario where control valve or block valve inadvertently close and hot process fluid continue to heat the CW in the heat exchangers. This scenario may include the initial Liquid expansion follow by flashing liquid passing the PSV


The design scenario is crucial for the PSV.
If you absolutely sure the PSV is provided for ambient heating thermal expansion, then you may consider to remove for maintainance / inspection. Yet you must have procedure in place i.e. partial drain CW to provide vapor space in the event of system short term shutdown. If it is for other scenarios and you operate the plant without this PSV, you probably put your plant in DANGER mode.

[Arvind Iyer]

Dear Friends,

Please refer to the attachment for the exact set-up.

To summarise the question for clarity: We want to assess if the PSV-210 on the CWR line can be
given to maintenance in a running plant.

Some engineers are of the opinion that these PSVs can be given to maintenance in a running plant
without any risk (and without any need to bypass the exchanger from service).

I am insisting that it cannot be given in a running plant and that it is a turn-around item.

What is your advice? Thanks.

Best Regards,
Arvind

Attached Files

•  PSV_CWR_LINE.JPG   61.11KB   150 downloads

[ankur2061]

Dear Friends,

Please refer to the attachment for the exact set-up.

To summarise the question for clarity: We want to assess if the PSV-210 on the CWR line can be
given to maintenance in a running plant.

Some engineers are of the opinion that these PSVs can be given to maintenance in a running plant
without any risk (and without any need to bypass the exchanger from service).

I am insisting that it cannot be given in a running plant and that it is a turn-around item.

What is your advice? Thanks.

Best Regards,
Arvind

Arvind,

It is obvoius from the attached sketch that this is a TERV or TSV application we are talking about in Cooling Water service. As I had mentioned earlier in my initial response this is not required to be a turn-down item. All you need to do is connect a nipple and hose to the valve upstream of the TERV as shown in the sketch and crack that valve open.

By cracking the valve open you will allow some water bleed which will prevent any pressure build-up due to thermal expansion. The bleed quantity can be adjusted to a very minimum and disposed off in your sewer till the time the TERV is back on-line. Obviously, operating procedures will also indicate that a tag needs to be provided that the TERV is under repair/maintenance and the butterfly valve in the return line as shown in your sketch will also be tagged with the sign "DO NOT CLOSE" or "LOCKED OPEN". A combination of operating procedures and maintenance ingenuity will ensure that you do not have to wait for a turn-down for your TERV.

Regards,
Ankur.

[JoeWong]

Arvind

Thermal Expansion (solar heating not process heating)
Look at the PSV-210 size (3/4" x 1"), it looks like a Thermal expansion relief valve (solar heating only).

Tube rupture
Looking at PSV-210 set pressure of 1030 kPag, i assumed the the design pressure of tube side is 1030 kPag (please confirm). Can you please advise the design pressure and temperature of both shell and tube side and/or test pressure of shell side ? Just wanna to see if tube rupture is a credible scenario.

Blocked-in-heat-on
The CW supply and return block valve (butterfly) are not locked open. There is potential of operator inadvertently close these valves and leave the hot processing fluid passing the heat exchanger (this is normally called Blocked-in-heat-on). Continuous heat supply from process fluid would lead to stagnant liquid in the CW side expand (liquid expansion not the simple thermal expansion due solar heating). Continuous heat input will heat the CW to boiling point and potential two phase relief. What is the maximum possible process fluid temperature or design temperature of process side ?

I just doubt if a simple 3/4" x 1" is sufficient. There is a potential flaw in this arrangement.
Besides, if above considerations (not only a thermal expansion relief valve) are remain unclear, then there are more reasons you shall not remove the PSV while plant still online as proposed.

[ankur2061]

Arvind

Thermal Expansion (solar heating not process heating)
Look at the PSV-210 size (3/4" x 1"), it looks like a Thermal expansion relief valve (solar heating only).

Tube rupture
Looking at PSV-210 set pressure of 1030 kPag, i assumed the the design pressure of tube side is 1030 kPag (please confirm). Can you please advise the design pressure and temperature of both shell and tube side and/or test pressure of shell side ? Just wanna to see if tube rupture is a credible scenario.

Blocked-in-heat-on
The CW supply and return block valve (butterfly) are not locked open. There is potential of operator inadvertently close these valves and leave the hot processing fluid passing the heat exchanger (this is normally called Blocked-in-heat-on). Continuous heat supply from process fluid would lead to stagnant liquid in the CW side expand (liquid expansion not the simple thermal expansion due solar heating). Continuous heat input will heat the CW to boiling point and potential two phase relief. What is the maximum possible process fluid temperature or design temperature of process side ?

I just doubt if a simple 3/4" x 1" is sufficient. There is a potential flaw in this arrangement.
Besides, if above considerations (not only a thermal expansion relief valve) are remain unclear, then there are more reasons you shall not remove the PSV while plant still online as proposed.

Joe,

IMO, this is a simple TERV employed in CW return circuit in heat exchangers in a petrochemical plant. This is not an Oil & Gas application and solar radiation is not a case to be considered because the heat exchanger is in-plant and not exposed to sun. I have dealt with practically dozens of S&T heat exchangers of the type similar to shown in the sketch during my operational career and the only case to be considered is the thermal expansion due to blocked-in water circuit. The TERV size is 3/4" X 1" typically for all cases that I have encountered. The CW circuit is a closed loop circuit with normal CW supply temperature of 30-32°C and CW return temperature of 40-42°C. The CW return header pressure is only a function of the frictional losses and static head considerations for the CW return circuit. The main CW header goes back to the cooling tower as an open circuit. Typically the design temperature of the CW piping is 60-70°C

The method I have described to take the TERV out of service with the HE still in-line is something I have practiced personally without any glitch.

I think we shouldn't be looking for ghosts where there are none.

Regards,
Ankur.

[JoeWong]

Ankur,
I respects your view point and your experience...

Arvind is here looks for advices and opinions. We have taken our effort to provide advices and opinions. Let Arvind to think about it and take the necessary decision.

[Arvind Iyer]

Dear Ankur,

Thanks for the inputs.

Dear JoeWong,

For your questions, please find the details in the attachment.

Best Regards,
Arvind

Attached Files

•  E_3005_design.JPG   24.04KB   72 downloads

[fallah]

A very small amount of bleed can be continuously maintained from the CW return circuit to prevent the scenario of thermal expansion if and when your CW TERV requires maintenance and hence dual TERVs are not required.

Because of not using dual TSVs to perform the suggested procedure,and typically in the case of TSVs application,a by-pass line shall be considered for maintenance requirements.

[JoeWong]

Sorry for late response...

The tube rupture to CW side may not be credible...

I have quick check on the liquid thermal expansion, the risk is pretty low based on my assumptions (but you have to calculate to confirm it). Remember, the liquid passing is possibly flashing liquid passing the PRV. So read API Std 521, app D.