Bypass relief in a heat exchanger 1

[MrBTU]

Hello,
We are designing over-pressure protection in an oil cooler, shell and tube heat exchanger to protect against high pressure drop during cold start up. This is to protect against damage to the baffles in the heat exchanger by supplying a relief valve around each baffle. We estimate that our baffles can withstand a 45 psi differential without failing, so we were planning a pipe loop around the baffle to the next with a 45 psi opening pressure. My question is - how do I add up the pressure drops in series, and account for the normal flow within the heat exchanger? For instance, if we test this unit, we will run the flow up (150 ssu lube oil) until we get 45 psi diff across the first baffle. At this point, the first bypass valve will pop, bypassing the flow to the second crosspass of the unit. We have 5 crosspasses in the heat exchanger (3 central cross passes and one under each nozzle). Will they all pop at once, or will it take an incremental increase in flow to get each additional one to pop? What will the flow vs pressure drop curve look like? If it was just one relief valve, it would be a linear rise and then flatten out after the valve opened, but what about three in series?
Thanks!

 

[BigInch]

It would probably help alot if you could manage to post a simplified diagram of the configuration. You've already lost me as to what pieces are in parallel and which are in series.

**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[MrBTU]

Here's a sketch..

 

[Chance17]

This design seems to need some impractically large safety valves to accomplish your objectives.

Assuming a liquid and near constant viscosity, there is a flow rate that represents 45-psi delta p across a single baffle. Because the baffle spacings and baffle area window are approx equal, the same flow would generate 45-psi delta P across all baffels.

A single control valve across the entire shell set to open for > 45-psi would be my recommendation.

 

[MrBTU]

Thank you Yitbos.
This is a relatively small heat exchanger and the bypass relief is actually in the baffles, so the external is only a representation. You've answered my question. Pressure drop is a function of flow rate, so the flow that pops the first one will pop them all.

 

[BigInch]

Then wouldn't the relief valves have to be dP activated and the maximum total pressure drop measured on dP be = 7 * 45 = 315 psi?

I don't understand why the upper baffles don't have similar relief valves, but maybe that isn't important.

Yes I agree with Yitbos, the same flow would tend to cascade all relief valves. After cascade the slope of your graph above would correspond to the K factor of a valve, provided that all other flow through the baffels was blocked. Otherwise there would be a slope representing the combination in parallel of any dP from flow still going through the baffles with the dP predicted by the K factor of a valve.







**********************
"The problem isn't working out the equation,
its finding the answer to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[rmw]

Why don't you just relieve between the inlet and outlet where you have the DP gage shown. After all, your convoluted multi valve path with all the reliefs that you show is essentially a path from the inlet to the outlet. Why make it so complicated?

rmw

 

[hydtools]

When we used air-to-oil coolers they were fitted with a bypass line that included a check valve that was designed to open at some pressure, in your case 45 psi. It was there for cold oil bypass. Seems like this would be simpler than using multiple relief valves.

Ted

 

[itsmoked]

simpler than using multiple relief valves

and more reliable.

Keith Cress
kcress -

http://www.flaminsystems.com