Design Considerations for Shell and Tube Heat

Exchangers

More on Shell and Tube Exchangers from Cheresources.com:

FREE Resources

Article: Basics of VaporizationQuestions and Answers: Heat TransferExperienced-Based Rules for Heat Exchangers and EvaporatorsChE Links: Search for "Heat Transfer"Students: Ask a Question in our ForumsProfessionals: Ask a Question in our Forums

Purchase / Subscription Resources

Book: Available TitlesOnline Store: Shell and Tube Exchanger Specification SheetOnline Store: Heat Transfer Category

    When preparing to design a heat exchanger, do you ever wonder where to start?  You've done

it before, but you hate that feeling of getting half way through the design and realizing that you forgot to consider one important element.  The thought process involved is just as important as the calculations involved.  Let's try to map out a heat exchanger design strategy.  We'll do so with a series of questions followed by information to help you answer the questions.

1.  Is there a phase change involved in my system?    A quick look at the boiling points compared with the entrance and exit temperatures will help you answer this question.

2.  How many "zones" are involved in my system?    "Zones" can best be defined as regimes of phase changes where the overall heat transfer

coefficient (Uo) will vary.  Using T-Q (Temperature-Heat) diagrams are the best way to pinpoint zones.  The system is defined as co-current or countercurrent and the diagram is constructed.  The diagram on the left illustrates the use of T-Q diagrams. These diagrams should accompany your basic (input-output) diagram of the heat exchanger.   Chemical #1 enters the shell at 200 0C as a superheated vapor. In Zone 1, it releases heat to the tubeside chemical (Chemical #2).  Zone 1 ends just a Chemical #1 begins to condense.  The tubeside (Chemical #2) enters as a liquid or gas and does not change phase throughout the exchanger.  Chemical #1 leaves Zone 1 and enters Zone 2 at its boiling temperature, Tb1.  T* marks the temperature of Chemical #2 when Chemical #1 begins to condense.  In Zone 2, Chemical #1 condenses to completion while Chemical #2 continues to increase in temperature.  The temperature of Chemical #2 when Chemical #1 is fully condensed is denoted at T**.  Finally, in Zone 3, both chemicals are liquids.  Chemical #1 is simply liberating heat to Chemical #2 as it becomes a subcooled liquid and exits the shell at 100 0C.     Defining zones is one of the most important aspects of heat exchanger design.  It is also important to remember that if your process simulator does not support zoned analysis (such as Chemcad III), you should model each zone with a separate heat exchanger.  Thus, the previous illustration would require 3 heat exchangers in the simulation.  BUT, do not draw 3 exchangers on your PFD (Process Flow Diagram).  This is all happening in one exchanger.

3.  What are the flowrates and operating pressures involved in my system?    This information is critical in establishing the mass and energy balance around the exchanger.  Operating pressures are particularly important for gases as their physical properties vary greatly with pressure.

4.  What are the physical properties of the streams involved?    If you're using a process simulator, obtaining the physical properties of your streams should be just a click of the mouse away.  However, if performing the calculation by hand, you may have to do some estimating as the streams may not be of pure substances.  Also, you should get the physical properties for each zone separately to ensure accuracy, but in some cases it is acceptable to use an average value.  This would be true of Chemical #2 in the tubes since it is not changing phase or undergoing a truly significant temperature change (over 1000C).  Physical properties that you will want to collect for each phase of each stream will include:  heat capacity, viscosity, thermal conductivity, density, and latent heat (for phase changes).  These are in addition to the boiling points of the streams at their respective pressures.

5.  What are the allowable pressure drops and velocities in the exchanger?    Pressure drops are very important in exchanger design (especially for gases).  As the pressure drops, so does viscosity and the fluids ability to transfer heat.  Therefore, the pressure drop and velocities must be limited.  The velocity is directly proportional to the heat transfer coefficient which is motivation to keep it high, while erosion and material limits are motivation to keep the velocity low.  Typical liquid velocities are 1-3 m/s (3-10 ft/s).  Typical gas velocities are 15-30 m/s (50-100 ft/s).  Typical pressure drops are 30-60 kPa (5-8 psi) on the tubeside and 20-30 kPa (3-5 psi) on the shellside.

6.  What is the heat duty of the system?    This can be answered by a simple energy balance from one of the streams.

7.  What is the estimated area of the exchanger?    Unfortunately, this is where the real fun begins in heat exchanger design!  You'll need to find estimates for the heat transfer coefficients that you'll be dealing with.  These can be found in most textbooks dedicated to the subject or in Perry's Chemical Engineers' Handbook.  Once you've estimated the overall heat transfer coefficient, use the equation Q=UoADTlm to get your preliminary area estimate.  Remember to use the above equation to get an area for each zone, then add them together.

8.  What geometric configuration is right for my exchanger?    Now that you have an area estimate, it's time to find a geometry that meets your needs.  Once you've selected a shell diameter, tubesheet layout, baffle and tube spacing, etc., it's time to check your velocity and pressure drop requirements to see if they're being met.  Experienced designers will usually combine these steps and actually obtain a tube size that meets the velocity and pressure drop requirements and then proceed.  Some guidelines may be as follows:  3/4 in. and 1.0 in. diameter tubes are the most popular and smaller sizes should only be used for exchangers needing less than 30 m2 of area.  If your pressure drop requirements are low, avoid using four or more tube passes as this will drastically increase your pressure drop.  Once you have a geometry selected that meets all of your needs, it's on to step #9.

9.  Now that I have a geometry in mind, what is the actual overall heat transfer coefficient?    This is where you'll spend much of your time in designing a heat exchanger.  Although many textbooks show Nu=0.027(NRE)0.8(NPR)0.33 as the "fundamental equation for turbulent flow heat transfer", what they sometimes fail to tell you is that the exponents can vary widely for different situations.   For example, condensation in the shell has different exponents than condensation in the tubes.  Use this fundamental equation if you must, but you should consult a good resource for accurate equations.  I highly recommend the following:  Handbook of Chemical Engineering Calculations, 2nd Ed., by Nicholas P. Chopey from McGraw-Hill publishers (ISBN 0070110212).  Also, don't forget to include the transfer coefficient across the tube wall and the fouling coefficient.  These can be very significant!

10.  What is the actual area of the exchanger using the 'actual' heat transfer coefficient?    If you recall, you used estimated heat transfer coefficients to get an initial area.  Now it's time to recalculate the area.

THE LOOP   Now you're on your way, pick a new geometry corresponding to your new ("actual") area, check the velocity and pressure drop, calculate the overall heat transfer coefficient again.  How does it compare with the previously calculated value?  If it is not within 5-10%, recalculate the process over and over (using your new value for Uo) until it does!  Sounds like alot of work.  Add in the fact that some of the individual heat transfer

coefficients require iterative solutions and it's not hard to see why people usually use a complex spreadsheet or a program to do this.  You can save some time by using estimates that you've undoubtedly seen, however you must realize that each time you estimate, you're losing accuracy.      Remember two main items:    1.  ZONED ANALYSIS2.   ACCURACY OF INITIAL OVERALL HEAT TRANSFER COEFFICIENTThe zoned analysis is the key to starting the process correctly.  The accuracy of the initial overall heat transfer coefficient will in part determine how many time you will be going through the calculation.

Other Considerations:

Materials of Construction Ease of Maintenance

Cost of Exchanger

Overall Heat Integration

Valve Sizing and Selection

More on Valves from Cheresources.com:

FREE Resources

Article: Using Equivalent Lengths of ValvesArticle: Rupture Pin Valve TechnologyChE Links: Search for "Valves"Students: Ask a Question in our ForumsProfessionals: Ask a Question in our Forums

Purchase / Subscription Resources

Book: Valve Selection HandbookOnline Store: Control Valve Flow DeterminationOnline Store: Control Valve Specification Sheet

    Sizing flow valves is a science with many rules of thumb that few people agree on.  In this article I'll try to define a more standard procedure for sizing a valve as well as helping to select the appropriate type of valve.  **Please note that the correlation within this article are for turbulent flow

STEP #1: Define the system    The system is pumping water from one tank to another through a piping system with a total pressure drop of 150 psi.  The fluid is water at 70 0F.   Design (maximum) flowrate of 150 gpm, operating flowrate of 110 gpm, and a minimum flowrate of 25 gpm.  The pipe diameter is 3 inches.  At 70 0F, water has a specific gravity of 1.0.

Key Variables:  Total pressure drop, design flow, operating flow, minimum flow, pipe diameter, specific gravity

STEP #2:  Define a maximum allowable pressure drop for the valveWhen defining the allowable pressure drop across the valve, you should first investigate the pump.  What is its maximum available head?  Remember that the system pressure drop is limited by the pump.  Essentially the Net Positive Suction Head Available (NPSHA) minus the Net Positive Suction Head Required (NPSHR) is the maximum available pressure drop for the valve to use and this must not be exceeded or another pump will be needed.  It's important to remember the trade off, larger pressure drops increase the pumping cost (operating) and smaller pressure drops increase the valve cost because a larger valve is required (capital cost).  The usual rule of thumb is that a valve should be designed to use 10-15% of the total pressure drop or 10 psi, whichever is greater.  For our system, 10% of the total pressure drop is 15 psi which is what we'll use as our allowable pressure drop when the valve is wide open (the pump is our system is easily capable of the additional pressure drop).

STEP #3:  Calculate the valve characteristic

For our system,

At this point, some people would be tempted to go to the valve charts or characteristic curves and select a valve.  Don't make this mistake, instead, proceed to Step #4!

STEP #4:  Preliminary valve selectionDon't make the mistake of trying to match a valve with your calculated Cv value.   The Cv value should be used as a guide in the valve selection, not a hard and fast rule.  Some other considerations are:a.  Never use a valve that is less than half the pipe sizeb.  Avoid using the lower 10% and upper 20% of the valve stroke.  The valve is much easier to control in the 10-80% stroke range.

Before a valve can be selected, you have to decide what type of valve will be used (See the list of valve types later in this article).  For our case, we'll assume we're using an equal percentage, globe valve (equal percentage will be explained later).  The valve chart for this type of valve is shown below.   This is a typical chart that will be supplied by the manufacturer (as a matter of fact, it was!)

For our case, it appears the 2 inch valve will work well for our Cv value at about 80-85% of the stroke range.  Notice that we're not trying to squeeze our Cv into the 1 1/2 valve which would need to be at 100% stroke to handle our maximum flow.   If this valve were used, two consequences would be experienced:  the pressure drop would be a little higher than 15 psi at our design (max) flow and the valve would be difficult to control at maximum flow.  Also, there would be no room for error with this valve, but the valve we've chosen will allow for flow surges beyond the 150 gpm range with severe headaches!    So we've selected a valve...but are we ready to order?  Not yet, there are still some characteristics to consider.

STEP #5:  Check the Cv and stroke percentage at the minimum flow   If the stroke percentage falls below 10% at our minimum flow, a smaller valve may have to be used in some cases.  Judgements plays role in many cases.  For example, is your system more likely to operate closer to the maximum flowrates more often than the minimum flowrates?  Or is it more likely to operate near the minimum flowrate for extended periods of time.  It's difficult to find the perfect valve, but you should find one that operates well most of the time.  Let's check the valve we've selected for our system:

Referring back to our valve chart, we see that a Cv of 6.5 would correspond to a stroke percentage of around 35-40% which is certainly acceptable.  Notice that we used the maximum pressure drop of 15 psi once again in our calculation.  Although the pressure drop across the valve will be lower at smaller flowrates, using the maximum value gives us a "worst case" scenario.  If our Cv at the minimum flow would have been around 1.5, there would not really be a problem because the valve has a Cv of 1.66 at 10% stroke and since we use the maximum pressure drop, our estimate is conservative.   Essentially, at lower pressure drops, Cv would only increase which in this case would be advantageous.

STEP #6:  Check the gain across applicable flowrates   Gain is defined as:

Now, at our three flowrates:Qmin = 25 gpmQop = 110 gpmQdes = 150 gpmwe have corresponding Cv values of 6.5, 28, and 39.  The corresponding stroke percentages are 35%, 73%, and 85% respectively.  Now we construct the following table:

Flow (gpm)

Stroke (%)

Change in flow (gpm)

Change in Stroke (%)

25 35110-25 = 85 73-35 = 38

110 73

150 85150-110 = 40 85-73 = 12

 

Gain #1 = 85/38 = 2.2Gain #2 = 40/12 = 3.3

The difference between these values should be less than 50% of the higher value.0.5 (3.3) = 1.65and 3.3 - 2.2 = 1.10.  Since 1.10 is less than 1.65, there should be no problem in controlling the valve.  Also note that the gain should never be less than 0.50.  So for our case, I believe our selected valve will do nicely!

OTHER NOTES:Another valve characteristic that can be examined is called the choked flow.  The relation uses the FL value found on the valve chart.  I recommend checking the choked flow for vastly different maximum and minimum flowrates.  For example if the difference between the maximum and minimum flows is above 90% of the maximum flow, you may want to check the choked flow.   Usually, the rule of thumb for determining the maximum pressure drop across the valve also helps to avoid choking flow.

SELECTING A VALVE TYPE

When speaking of valves, it's easy to get lost in the terminology.  Valve types are used to describe the mechanical characteristics and geometry (Ex/ gate, ball, globe valves).  We'll use valve control to refer to how the valve travel or stroke (openness) relates to the flow:1.  Equal Percentage:  equal increments of valve travel produce an equal percentage in flow change2.  Linear:  valve travel is directly proportional to the valve stoke3.  Quick opening:  large increase in flow with a small change in valve stroke

So how do you decide which valve control to use?  Here are some rules of thumb for each one:1.  Equal Percentage (most commonly used valve control)a.  Used in processes where large changes in pressure drop are expectedb.  Used in processes where a small percentage of the total pressure drop is permitted by the valvec.  Used in temperature and pressure control loops

2.  Lineara.  Used in liquid level or flow loopsb.  Used in systems where the pressure drop across the valve is expected to remain fairly constant (ie. steady state systems)

3.  Quick Openinga.  Used for frequent on-off serviceb.  Used for processes where "instantly" large flow is needed (ie. safety systems or cooling water systems)

Now that we've covered the various types of valve control, we'll take a look at the most common valve types.

Gate ValvesBest Suited Control:  Quick Opening

Recommended Uses:1.  Fully open/closed, non-throttling2.  Infrequent operation3.  Minimal fluid trapping in line

Applications:  Oil, gas, air, slurries, heavy liquids, steam, noncondensing gases, and corrosive liquids

Advantages:                             Disadvantages:1.  High capacity                         1.   Poor control2.  Tight shutoff                           2.   Cavitate at low pressure drops3.  Low cost                               3.   Cannot be used for throttling

4.  Little resistance to flow

 

Globe ValvesBest Suited Control:  Linear and Equal percentage

Recommended Uses:1.  Throttling service/flow regulation2.  Frequent operation

Applications:  Liquids, vapors, gases, corrosive substances, slurries

Advantages:                         Disadvantages:1.  Efficient throttling               1.   High pressure drop2.  Accurate flow control        2.  More expensive than other valves3.  Available in multiple ports

 

 

Ball ValvesBest Suited Control: Quick opening, linear

Recommended Uses:1.  Fully open/closed, limited-throttling2.  Higher temperature fluids

Applications: Most liquids, high temperatures, slurries

 

 

Advantages:                         Disadvantages:1.  Low cost                           1.   Poor throttling characteristics2.  High capacity                     2.   Prone to cavitation3.  Low leakage and maint.4.  Tight sealing with low torque 

Butterfly ValvesBest Suited Control:  Linear, Equal percentage

Recommended Uses:  1.  Fully open/closed or throttling services2.  Frequent operation3.  Minimal fluid trapping in line

Applications:  Liquids, gases, slurries, liquids with suspended solids

Advantages:                        Disadvantages:1.  Low cost and maint.        1.  High torque required for control

2.  High capacity                  2.   Prone to cavitation at lower flows3.  Good flow control4.  Low pressure drop

 

Other Valves    Another type of valve commonly used in conjunction with other valves is called a check valve.  Check valves are designed to restrict the flow to one direction.  If the flow reverses direction, the check valve closes.   Relief valves are used to regulate the operating pressure of incompressible flow.  Safety valves are used to release excess pressure in gases or compressible fluids.

References:

Rosaler, Robert C., Standard Handbook of Plant Engineering, McGraw-Hill, New York, 1995, pages 10-110 through 10-122Purcell, Michael K., "Easily Select and Size Control Valves", Chemical Engineering Progress, March