Designing Shell-and-Tube Exchangers

Will this exchanger do the duty? Developing a design envelope. Choosing the best design.

Hyprotech 2002

Lecturers Guide
Copyright Hyprotech UK Ltd holds the copyright to these lectures. Lecturers have permission to use the slides and other documents in their lectures and in handouts to students provided that they give full acknowledgement to Hyprotech. The information must not be incorporated into any publication without the written permission of Hyprotech.

Lecture series
Introduction to heat exchangers Selection of the best type for a given application Selection of right shell and tube Design of shell and tube
Hyprotech 2002

Lecturers Guide

Contents
Overview of design Single phase rating methods used in DEVIZE
Tube side Shell side

Design envelope concept in DEVIZE Demonstration of DEVIZE

Hyprotech 2002

Lecturers Guide

DEVIZE
DEsign VIsualiZE Program which uses visual features to guide you through the design process

DEVIZE

Hyprotech 2002

Lecturers Guide

Overall design process

Decide fluid allocation (last lecture) Decide TEMA type (last lecture) Make some guesses about the design Generate design envelope using DEVIZE Do a rating using DEVIZE Review the envelope and the rating in order to improve your initial guess Finally select the best design
Lecturers Guide

Hyprotech 2002

Repeat until satisfied with design

Starting point of rating

We know The duty which we need to achieve (i.e. flow rates and temperature changes of the two streams) The full exchanger geometry The allowable pressure drops for the two streams We are checking whether that exchanger can do that duty within the imposed constraints
Hyprotech 2002

Lecturers Guide

The thermal rating

The actual heat transfer area (based on the tube o.d.), Ao is known We calculate the required area, Areq from

Areq =

Q U o Tm

One output of the thermal rating is the ratio

RA =
Hyprotech 2002

Ao Areq

Lecturers Guide

yw

Overall coefficient
yw Thot Di Do

Tcold

We have thermal resistances in series

D y D 1 1D 1 o = +ri o + w o +ro + Uo i D D w Dw o i i
Where
Hyprotech 2002

Dw

1 ( Do + Di ) 2

Lecturers Guide

Thermal conductivity
Typical values of thermal conductivity
M aterial Stainless Steel Copper Alum inium Carbon Steel G ases Liquids Polystyrene foam Pipe Lagging
Hyprotech 2002

W /m k 15 390 208 50 0.02 - 0.3 0.03 - 0.7 0.003 0.092

Lecturers Guide

Typical values of stream coefficients

Fluid Water Water Steam Organic Organic Organic Gas Gas
Hyprotech 2002

State Single Phase Boiling <5Bar Condensing 1 Bar Single Phase 0.5-2.5 cp Boiling 0.5-2.5 cp Condensing 0.5-2.5 cp 1 Bar 10 Bar

htc W/m2.k 5000 - 7500 3000 - 10000 10000 - 15000 750 -1500 1000 - 3500 1500 - 4000 80 - 125 250 - 400

Lecturers Guide

Typical fouling resistance

Fluid Water Water Steam Organic Organic Organic Gas Gas
Hyprotech 2002

State Single Phase Boiling <5Bar Condensing 1 Bar Single Phase 0.5-2.5 cp Boiling 0.5-2.5 cp Condensing 0.5-2.5 cp 1 Bar 10 Bar

r (m2.k/W) 0.0001 - 0.00025 0.0001 - 0.0002 0 - 0.0001 0.0002 - 0.001 0.0001 - 0.0003 0.0001 - 0.0003 0 - 0.0001 0 - 0.0001

Lecturers Guide

Single phase exchangers

We will now concentrate on a single phase in both streams This is a limitation of DEVIZE But the principles of the design process are similar for boiling and condensing streams In single phase
coefficients do not vary much and can be treated as constant and Tm = FT TLM
Hyprotech 2002

Lecturers Guide

Tube side heat transfer coefficient

Nu = 0.023 Re 0.8 Pr 0.4 ,Re 8000 Re < 2000 Nu = 3.66, Gz 9 Nu = 1.75Gz1/ 3 , Gz > 9 d Gz = Re Pr i 4 L

In the transition region, Nu is calculated from a linear interpolation with Re of the values at Re = 2000 and 8000
Note: wall-to-bulk property variations and natural convection effects are neglected
Hyprotech 2002

Lecturers Guide

Tube side pressure drop

General equation
L u2 p = 4 f di 2

Laminar flow (transition taken as 16 crossover with f = ,Re 837 Re turbulent equation) Turbulent flow in a commercial rough tube
Hyprotech 2002

f = 0.0035 +

0.264 ,Re > 837 Re0.42

Lecturers Guide

Nozzle, tube entry and return losses

u 2 p = K 2
K is taken as 1.8 per tube-side pass to allow for tube entry, tube exit and header losses 10% of the total pressure drop is assumed to occur in the nozzles

Hyprotech 2002

Lecturers Guide

Shell side - the Bell Method

DEVIZE uses the Bell-Delaware method as set out by Taborek in HEDH This method starts with the coefficients and pressure drops for ideal cross flow and then corrects these for the non-idealities which occur in real shell-side flows The method gives reasonable accuracy while remaining simple More accurate proprietary methods have been developed based on network models
Hyprotech 2002

Lecturers Guide
HEDH is the Heat Exchanger Design Handbook available from Ray Johnson HEDU office 3 St. Peters Street, Wallingford, OX10 0BQ, UK Tel: +44 (0)1491 834930

Shell-side non-idealties
Window effects

Tube-to-baffle leakage

Bypass
Hyprotech 2002

Shell-to-baffle leakage

Lecturers Guide

Bell method for heat transfer = Ideal JW J L JB J Lam

Where Jw corrects for some tubes being in the window JL corrects for leakage through and around the baffle JB corrects for flow around the bundle JLam corrects for special effects which occur at very low Reynolds numbers (not applied in DEVIZE)
Hyprotech 2002

Lecturers Guide

What do the factors depend upon?

Jw - ratio of tubes in cross flow to tubes in windows JL - ratio of leakage areas to cross-flow area JB - ratio of bypass flow area to cross flow area and the number of pairs of sealing strips

Sealing strips to reduce bypass

Hyprotech 2002

Lecturers Guide

Example correction factor - bypass

Hyprotech 2002

Lecturers Guide

Ideal crossflow coefficient

Ideal = uCp j Pr 2 /3
The j-factor is given as a function of Reynolds number, defined by Re = umin di / Where umin is the velocity calculated at the minimum flow area near the equator of the bundle (assuming no leakage and bypass)
Hyprotech 2002

Lecturers Guide

Shell-side pressure drop and vibration

Pressure drop The Bell method for pressure drop is similar to that for heat transfer but with some extra complications in the end zones An additional 10% is assumed for the nozzles Vibration DEVIZE uses the methods in ESDU Data Item 87019 for fluid-elastic instability
Hyprotech 2002

Lecturers Guide
ESDU 27 Corsham Street London, N1 6UA esdu@esdu.com

Outputs of rating
Area ratio (already discussed): RA = Ao / Areq Pressure drop ratios for the shell side and the tube side: Rp= pcalc/ pspec Velocity ratio for vibration ratio: Rv = u / ucrit For an acceptable design, RA > 1, RpT < 1, RpS < 1, Rv < 1 The closer these ratios are to 1, the tighter the design (i.e. the lower the safety margins)
Hyprotech 2002

Lecturers Guide

Thermal resistance diagram

Good way of viewing where the main resistances lie and therefore where it is best to make changes to improve the design

y D D 1D 1 1 = +ro + w o +ri o + o Uo o w Dw D i i D i
Shell Side Tube Side

Stream
Hyprotech 2002

Fouling

Wall

Lecturers Guide
Heat transfer engineers talk in terms of heat transfer coefficients for streams but thermal resistances for fouling. It is helpful to put this all on the same bases and the best way to do this is in terms of resistances as in the little diagram which is output in a DEVIZE rating. During the design, it is usually best to find ways of lowering the highest resistance. In addition to showing where the main resistances are, this illustrate the proportion of the heat exchanger surface area required to cope with each resistance. Often, this diagram shows dramatically that a large proportion of the surface area is required to overcome the fouling. In such cases, the choice of fouling resistance should be questioned. Another useful way of looking at this diagram is that each part is proportional to the temperature difference resulting from the thermal resistance. Hence, the diagram can be used a s quick guide to what the wall temperature is. This may be important if high wall temperatures must be avoided (to prevent scale formation, say).

Envelope design concept

Description of the envelope design concept and the demonstration of DEVIZE Starting point is to make guesses about key geometrical features
number of passes baffle pitch (say as factor on the shell diameter) baffle cut (say to equalise the cross-flow and window-flow areas)

Hyprotech 2002

Lecturers Guide

Tube side pressure drop

Shell diameter

For given shell diameter DEVIZE can calc. The tube length which just uses up the p

Tube length
Hyprotech 2002

Lecturers Guide
Take the case of a single pass, as an example, for a known shell diameter we can calculate the number of tubes that fit in the shell. Hence, we can calculate the velocity in the tubes and the pressure gradient along the tubes. From the gradient, we can calculate the exact tube length to just use up the allowably pressure drop.

Tube-side pressure drop limit

Repeating for range of shell diameters gives curve

Shell diameter

Designs valid in this region

Not valid in this region

Tube length
Hyprotech 2002

Lecturers Guide
Repeating the calculation for different shell diameters gives the locus shown by the blue curve. Remember, blue equals tube from now on in this lecture. The region for valid designs is shown yellow here. This convention is used again later in the lecture but is dropped for the next few slides.

Shell-side pressure drop limit

We can do similar thing for the shell side

Shell diameter

Shell-side p limit

Tube-side p limit

Tube length
Hyprotech 2002

Lecturers Guide
A similar set of calculations may be done for the shell side pressure drop thus giving the green curve. Immediately we see that the tube side pressure drop is no longer having an effect on the design in this case. Hence we might be able to improve the design by introducing tube side enhancement of increasing the number of tube side passes. We might not for other reasons but these curves are already giving us useful clues for improving the design.

Heat transfer limit

And then for the heat transfer Heat transfer area limit Shell diameter

drop lim Pressure

its

Tube length
Hyprotech 2002

Lecturers Guide
We can do the same thing for the heat transfer area thus giving the red curve.

Add limits on tube-side velocity

Shell diameter

Valid Valid

Min.

Max.

Tube length
Hyprotech 2002

Lecturers Guide
Often we set limits on the maximum and minimum velocities in the tubes. The minimum may be to avoid fouling while the maximum would be to avoid erosion.

Danger areas for vibration may also be shown

Shell diameter

Min.

Max. Vibration Tube length

Hyprotech 2002

Lecturers Guide

Highlight valid design region

Shell diameter

Design envelope

Min.

Max.

Tube length
Hyprotech 2002

Lecturers Guide

Rating at optimum point

Smallest diameter and shortest tubes so likely to be the cheapest Min.

Shell diameter

Max.

Tube length
Hyprotech 2002

Lecturers Guide
Doing a rating at this point will generate a resistance diagram which will give more clues on how to improve the design.

Preferred shell diameters and tube lengths

Shell diameter

Min. A B Max.

Tube length
Hyprotech 2002

Lecturers Guide
If you are working with fixed tube lengths and shell diameters, you can introduce the grid of these as shown and select the point closest to the optimum. This is point A but a good designer might think of ways of modifying the design so that point B falls within the yellow area.

Design process
Start with a set of assumptions about number of passes, baffle configuration, etc. Generate envelope Change initial assumptions to obtain better envelopes Then Rate good geometries within the best envelopes Consider advantages of ratings outside the envelope assuming constraint may be relaxed Select the best Rating
Hyprotech 2002

Lecturers Guide

Demonstration of DEVIZE

Hyprotech 2002

Lecturers Guide
All this will become clear by using DEVIZE and by producing a demonstration. There is an article in Chemical Technology Europe, July/August 1996 which takes you through a design example.