8 replies to this topic

[sfhh]

Hi

In Shell and tube sizing software (like HTRI), there is over-design%,(as I know it is U-dirty/U-service)
What does it mean?

When we size heat exchanger, should we take safety factor by
using "sizing flow rate" = normal flow rate x 1.1 or Can we just use the above said over-design%?

[latexman]

It is traditional on HX design to use U-dirty/U-service (or fouling factors or dirt factors, etc.). What does it mean? What you are doing is setting the future maintenance schedule for cleaning of the HX. If the time between cleanings is less than planned turnarounds, you may need multiple HXs or some other method to achieve planned turnaround time.

[Himanshu Sharma]

Normally I write two different over design margins while specifying a Heat exchanger datasheet

Margin on Flow :Hydraulic margin, i.e exchanger pressure drop shall remain in prescribed maximum delta p ,even if flow is increased up to the max flow level(Including design margin).

Margin on Duty:Area margin ,this translates to excess area that is required above the area requirement at normal duty.

[sheiko]

Hi,
This article from HTRI deals with HX design margins:
https://www.htri.net...407_Bennett.pdf

Edited by sheiko, 12 November 2011 - 09:08 PM.

[kkala]

The Udirty/Uservice margin may not be applied here (never heard of), but it could be used instead of fouling factors. Fouling factors are usually determined at the project launch for each service. When we prepare the heat exchanger data sheet, we select the proper fouling factor for each service from the engineering data book (whatever its name can be) already prepared and approved by Client.
Fouling factors may affect the exchanger size considerably, I remember a case of LPG vaporizer where clean service required about 1/3rd of area for dirty service (http://www.cheresour...h__1#entry35347).
Local practice here adds a 10% overdesign factor on exchanger surface, without modifying any parameter in exchanger data sheet. That is process requirements (duty, flows, temperatures, ΔP) are specified and Mechanical Dept runs b-jac to define exchanger characteristics. After fixing the exchanger and its area required A (in dirty service), the data sheet to bidders contain the condition that exchanger area should not be lower than 1.1*A.

[DB Shah]

Ud/Uservice is same as the excess area (1.1A ) termed in kkala post.
For eg, Q=10000, Ud = 50 & LMTD = 20 (MKS)
Area required = 10000/(50*20) = 10 m2, Now you provide area margin of 1.1 ie area provided = 10*1.1=11 m2.
Now recalculate U => U = 10000/(20 * 11) = 45.5. This U is termed as Uservice and hence
Ud/Uservice = 50/45.4545 = 1.1

[Chellani]

One question related to this.

As most of you've suggested that we provide area higher than what is required which means HE would deliver more duty than what is shown in heat and weight balance (HWB) sheet or what is required in plant and hence we have to provide bypass to get what we want.
Do you show flow in this bypass in PFD and HWB or it is responsibility of operaations/control system to adjust bypass based on actual requirement?

Hope I am not confusing everyone.

[Himanshu Sharma]

Kudos Chellani !

A great concern raised ,for once i have seen the exchangers in crude units choking up after a maintenance turndown as the startup was on 50% flow rate and there was so much excess area available for the duty(unfortunately other side was steam heating) that material in shell actually coked.

Normally this is not a problem as for both sides process fluid exchanging heat/cold because heat transfer is also governed by flow rate of streams (exchanging heat) and Temp approach available.

For Trim Coolers/Steam heaters a real precise temp control is not required for general refinery operations, so a few degrees here and there is accepted.In case a precise temp at exchanger outlet is required you have to employ a control either or utility side or process side (remember Tempered water cooler process control been employed in VDU for congealing services like MVGO & HVGO).

At the end of the day exchanger fouling plays its role and then this 10% margin we have no longer seems sufficient !

[ankur2061]

All,

Long time back there was a general discussion on overdesign or in that particular thread titled as "Conservatism in Chemical Process Design". Some great comments and observations were provided by the participants such as Andrei, Doug, Qalander and Joe Wong. Here is the link:

http://www.cheresour...h__1#entry23600

Regards,
Ankur.