Process Plant Accidents

Accidents happen....that is for sure. But when the stakes are as high as they are in the chemical industry, every effort must be made to ensure that diasters are avoided. One of the best ways to avoid accidents is to learn from others. Let's take a look at some past accidents that were all too real and derive a simple lesson from each one.

A Temporary Modification That Proved Deadly

Nypro Factory, Flixborough, United Kingdom in 1974 A series of six reactors were installed side by side, each installed a little lower than the previous one to allow for gravity flow between them. The reactors were joined by 28 in. diameter pipe with an expansion bellow on the end of each pipe.

The bellows were installed on the short pipe runs to allow for expansions and small movements that typically accompany operating conditions such as those at Nypo (150 ⁰C and 10 bar). These reactors were usually cooled via brine to jackets on the outside of the vessels. The brine system had to be taken offline for repairs so city water was used in the interim. Reactor #5 collapsed soon after, but no significant leak was experienced.

Error #1

The city water was supplied to the cooling jackets at 9 bar, a pressure higher than the brine system operated under. A simple material stress analysis may have been able to predict the material's limitations.

A 20 question modification approval form was completed before any modifications were made. The questions were not answered properly and the form was treated as a formality. A professionally qualified engineer was not available to review the suggested modification.

Error #2

Any plant modification should be taken seriously no matter how simple it seems. All modifications should be approved by qualified personnel.

While Reactor #5 was being repaired or replaced, the modification that was approved and installed is shown below:

Reactor #5 was replaced with a 20 in. pipe with two bends to accomodate the height difference. The existing bellows were left in place at the ends of the temporary pipe. The temporary pipe was supported by resting it on scaffolding.

Error #3

The third error, and most costly, was the lack of physical bracing or support for the pipe. The original pipe runs were shorter and supported firmly as to minimize the pipe movement and to allow the bellows to operate properly.

Result

The temporary pipe was allowed to twist and move laterally due to the lack of proper support. This is a textbook recipe for mechanical failure. That's exactly what happened in one of the bellows. Approximately 50 tons of cyclohexane was released, mixed with air, and exploded. 28 people lost their lives.

Lack of Concentration of Faulty Design

An operator was asked to maintain the temperature of a reactor at 60 ⁰C. The control panel was set up in such a manner that it may have resembled the diagram below:

Essentially, the set point should have been established by percentage of the temperature range from 0 to 200 ⁰C. It's not difficult to guess what happened. The operator simply set the instrument to 60, which corresponded to a reactor temperature of 120 ⁰C! The result was a runaway reaction, an overpressurized reactor, liquid release, and injured operators.

Error #1

The operator simply did not pay attention to what he/she was doing. The lack of concentration could have resulted in much more serious consequences.

Error #2

The controller lacked an intuitive operating "feel". The controller almost begged for an incident such as this to occur. Apparently, there was no temperature safety device installed on the unit...another easily avoidable error.

Understanding Why Testing is Being Done

A new storage tank was being filled with water to check for leaks and mechanical stability. While the testing was being performed, two welders were on top of the tank completing the hand rails for the tank.

Error #1

The tank was being tested for a good reason...to be sure it is safe. We cannot know that it is safe until after the testing. Therefore the welders were on a tank working during a potentially dangerous operation.

The line used to feed the water to the tank was previously used for transporting gasoline. Some residual gasoline entered the tank with the water. Once inside the tank, the gasoline immediately floated to the top. The welders accidentally ignited the gasoline vapor. Fortunately, neither welder was seriously injured.

Error #2

Since the water feed line previously contained gasoline, it should have been flushed briefly to a drain before being used to fill the tank.

Scaling Up Is Not Elementary

Scaling up of chemical processes has resulted in many accidents throughout history. Designers sometimes fail to realize that increasing the volume of a reactor does not mean that the surface area and heat transfer will increase proportionally. Consider a cylindrical reactor:

After scale up, the volume increased by about 2.7 times the original design. However, the surface area only increased by 1.9 times the original design. Consider an example where a reaction was studied in a laboratory. In the lab, there was no observed rise in temperature so the reaction was deemed "thermally neutral" (which is rare). When the reaction was moved to the pilot plant level, no cooling supply was added to the reactor.

Error

The heat loss through the laboratory glassware was 3-6 W/kg 0C. The reaction produced 2.0 W/kg 0C. Therefore no temperature change was noted until the reaction scale was increased.

A combination of an increased reactor size (2.5 m³) and a decreased heat loss rate (0.5 W/kg⁰C in the pilot plant) resulted in a runaway reaction. Fortunately, a relief valve was in place and a serious incident was avoided.

After reviewing these incidents, its easy to see how knowing what caused past accidents can prevent future accidents from happening. With that in mind, I would urge everyone to read the reference cited below. Mr. Trevor Kletz assembled a well-written compilation of process plant accidents in his book. It's a fantastic resource!

Methanol Plant Capacity Enhancement

Brief History - Enhancing the Methanol-I Plant Capacity

Originally, this plant was designed to operate on feed gas from an ammonia plant consisting of a gas mixture of 75% hydrogen, 22% carbon dioxide, 1% carbon monoxide and some inerts. The reaction of the methanol in gas rich in CO₂ is milder as it produces water along with methanol. The crude methanol concentration is also lower. Water further retards the rate of reaction. The two reactions involved here are:

H₂ + CO₂ ----> CH₃OH + H₂O + 9.8 kcal/kgmol

H₂ + CO ----> CH₃OH + 21.6 kcal/kgmol

Normally, a gas mixture of H₂ + CO + CO₂ is used in a proportion measured in terms of a “R” value (H₂-CO₂)/(CO+CO₂) equal to 2.0 to get an optimum methanol conversion per pass.

Changes to the process included:

1. A methanol chiller was introduced in the gas cooling circuit at the reactor outlet to reduce the methanol concentration and temperature in the recycle gas which helped to increase the methanol production from 4~5 MTD up to 75 MTD level.

2. Setting up a synthesis gas generation unit (SGGU) to supply CO rich gas from natural gas reformer in February 1998. This gas composition is better for methanol production compared to the rectisol wash gas which is rich in CO₂. The synthesis gas and distillation loops were debottlenecked by replacing of some control valves, installation of exchangers, and other modifications. The capacity was boosted to 100 to 120 MTD.

3. Replacement of the refining column trays with high capacity Superfrac™ trays from Kostch Glistch India Ltd in October 2002. This, along with other peripheral modifications, were made to increase distillation capacity to 145 MTD.

4. Replacement of the quench adiabatic methanol convertor to Linde’s Isothermal Reactor and debottlenecking of the distillation loops for higher capacity. The capacity of the plant was increased to 160 MTD in September 2003.

Major advantages of Isothermal Reactor include:

Lower pressure drop in reactor
Less temperature variation
Increased life of catalyst
Narrow band of temperature differences in the reactor catalyst bed
Sustained Production Level throughout the catalyst life due to better conversion
Less by-product formation
Effective heat recovery

Figure 1 below shows the temperature profile in the isothermal reactor.

Figure 1: Comparing Reactor Temperature Profiles Before and After the Changes

Compared to the expected 160 MTD production capacity, the unit has achieved a stable production level of 185~190 MTD.

A flow diagram of the new loop is shown in Figure 2 below. In this article, we’ll focus on this latest dimension added to the plant, highlighting the re-commissioning experiences.

Figure 2: Changes in the Methanol Synthesis and Distillation Loops

Figure 3: Methanol Synthesis and Distillation Loops After Changes

Plant Re-commissioning with the Isothermal Reactor

Following the replacement of the quench reactor with the Isothermal reactor from Linde, the plant was ready for start up. The following details the activities associated with start up after the changes were made.

Basic and Detail Engineering - Design Fundamentals

The original plant was designed by Linde with process licensing from ICI. Linde performed the basic engineering for the loop modification and the detailed engineering for the new Isothermal reactor. Based on the data for the new design conditions, a debottlenecking study on the distillation section was carried out in-house by our Technical Services department. Major pre-fabrication work and in-plant erection of the loops which were to be replaced was completed before the final shutdown of the plant. A shutdown schedule of 11 days was planned.

Outline of the Pre commissioning activities

The piping loops were identified and broken down into various process loops per the P & IDs. The plant was broadly classified into three independent sections: synthesis loop, makeup gas loop, and distillation loop. This helped prioritize tasks such that the synthesis and related loops were made ready first. The loops, which were erected before shutdown, were prepared for commissioning by flushing / blowing. Based on the service, the plans for flushing / blowing were prepared and discussed with the mechanical and instrument groups to streamline the activities. All instruments in the circuit were removed from the lines.

The following procedures were used:

For gas lines: Gasket blowing with plant air was carried out starting from 1.0 barg up to 3.5 barg repeatedly, until there was no rust / dust in the line. This was followed by nitrogen passivation / drying.

For liquid lines: Air blowing followed by water flushing was carried out. This was followed by nitrogen passivation / drying.

For steam lines: Gradual warming of the header before insulation was applied for grease removal and rust flushing through the trap bypass. Then steam blowing at full capacity was carried out for half an hour by diverting the open end at a safe location. The header was allowed to cool. This cycle was repeated again till clear condensate was discharged in the trap bypass.

For Running Machines: There was a pair of process pumps in each service. One pump online and one spare. With the higher capacity, some pumps were replaced for higher capacity. The main crude feed pumps and refining column reflux pumps were replaced. With spare pumps, the plant operation was not interrupted during the pump changes. Each replacement took 12 days and included the modification of the base, pipeline, motor, and other ancillary pieces.

Likewise, four control valves were replaced via proper coordination between the operations and project teams. The prefabricated loops were also washed or blown and then dried with nitrogen. These were kept inert and sealed at their ends until they were to hooked up during the shutdown. This also helped reduce the pre-commissioning time for the plant. The start-up boiler feed water circulation pump was commissioned and stabilized prior to shutdown as soon as the errection of the reactor steam drum system was completed.

Both Methanol-I and SGGU operate independently. It was not necessary to shutdown SGGU for the commissioning of Isothermal reactor in the Methanol-I synthesis loop. The natural gas compressors in the SGGU plant get cooling water from the Methanol-I plant header. Since cooling tower was to be taken offline, temporary arrangements to supply an alternate water source was planned to keep the natural gas compressors in the SGGU running. This was implemented prior to shutdown, avoiding a stoppage of the SGGU plant.

Reactor Catalyst Charging

This was the first reactor of its kind at GNFC with spiral wound coils within the shell.

The catalyst is to be charged on the shell side, while the cooling medium (boiler feed water) flows in the tubes via thermosiphon. During the startup, the boiler feed water circulation pump maintains the water circulation.

As shown in Figures 4 and 5, the reactor coils are supported at both the ends by six support strips moving radially outward from the central mandrill. This divides the cross section of the reactor into six equal parts. These were taken as the basis for charging the catalyst and numbered from 1 to 6 inside the reactor.

Figure 4: Internal View of the Isothermal Reactor

Figure 5: Top Internal View of the Isothermal Reactor

Two catalyst charging nozzles were used with hoppers and 2 ½” dia flexible hoses for charging the catalyst - SACK WISE under the supervision of Linde. A table was prepared to log the number of bags charged per round and the subsequent dip achieved, which showed the packing uniformity. This proved to be a very successful method of charging with good packing density with less than 20 mm of variation in the final height adjustment. A flat and heavy plumb with strong cotton thread was used for taking the dip.

Approximately equal quantities of 20 kg balls/catalyst were filled in HDPE sacks before the start of loading. About 5.2 m³ alumina balls were filled first in four rounds of sack charging. The catalyst bed was leveled so that the balls were just inside the tube coiled bundle. The first dip of catalyst was taken after charging almost half of the catalyst. Thereafter, while monitoring the height, charging continued to completion over approximately two (2) days.

Commissioning activities

The synthesis loop was made available earlier than the distillation loop (6 ½ days) while the total shutdown period was compressed to 8 days by effective identification of the priority of each job. The effectiveness of the pre-commissioning activities was evident during post-commissioning. There were no plugged strainers, control valves by-passing, nor false signals during or after the startup of the plant. Re-commissioning of the plant was completed in less than 4 days time. While, the distillation section modifications were being completed, the synthesis loop pre-commissioning activities were completed. While the catalyst heat up and reduction was proceeding, crude methanol production was coming online.

Peak production levels for the plant were achieved while testing the plant at different feed gas mixtures. The plant has met all process guarantees. Of particular interest has been an improved yield of methanol due to a higher conversion rate and stable reaction conditions. Less by-product formation has led to a reduction of loading in the distillation section.

Conclusion

From this experience, we see that the plant capacity can be increased by understanding the basic principles of reaction kinetics and unit operations. Through integration of technology and the use of improved catalyst, this little plant had been transformed into a giant producer. Proper planning of critical activities like catalyst charging, pre-commissioning of loops, commissioning and guarantee test runs can ensure success.

Crystallization

Crystallization refers to the formation of solid crystals from a homogeneous solution. It is essentially a solid-liquid separation technique and a very important one at that.

Example of Crystallization

1. Water freezing

2. Removing sucrose from beet solutions

3. Removing KCl from an aqueous solution

Crystals are grown in many shapes, which are dependent upon downstream processing or final product requirements. Crystal shapes can include cubic, tetragonal, orthorhombic, hexagonal, monoclinic, triclinic, and trigonal. In order for crystallization to take place a solution must be "supersaturated". Supersaturation refers to a state in which the liquid (solvent) contains more dissolved solids (solute) than can ordinarily be accomodated at that temperature.

As with any separation method, equilibrium plays an important role. Below is a general solubility curve for a solid that forms hydrate (a compound that has one or more water molecules attached) as it cools.

In Figure 1, X may be any solid that can form hydrates such as Na₂S₂O₃. The number of hydrate molecules shown in Figure 1 is strictly arbitrary and will vary for each substance.

So how do you grow crystals? Let's consider an example that is fairly easy to envision. Take a pot of boiling water and add table salt while stirring to make a water-salt solution. Continue adding salt until no more salt will dissolve in the solution (this is a saturated solution). Now add one final teaspoon of salt. The salt that will not dissolve will help the first step in crystallization begin. This first step is called "nucleation" or primary nucleation. The salt resting at the bottom of the pot will provide a site for nucleation to occur.

Primary nucleation is the first step in crystallization. Simply defined, it's the growth of a new crystal

On an industrial scale, a large supersaturation driving force is necessary to initiate primary nucleation. The initiation of primary nucleation via this driving force is not fully understood which makes it difficult to model (experiments are the best guide). Usually, the instantaneous formation of many nuclei can be observed "crashing out" of the solution. You can think of the supersaturation driving force as being created by a combination of high solute concentration and rapid cooling. In the salt example, cooling will be gradual so we need to provide a "seed" for the crystals to grow on. In continuous crystallization, once primary nucleation has begun, the crystal size distribution begins to take shape. Think about our salty water, as you look at Figure 2 describing the progression of crystallization.

Figure 2: Progression of Crystallization

The second chief mechanism in crystallization is called secondary nucleation. In this phase of crystallization, crystal growth is initiated with contact. The contact can be between the solution and other crystals, a mixer blade, a pipe, a vessel wall, etc. This phase of crystallization occurs at lower supersaturation (than primary nucleation) where crystal growth is optimal.

Secondary nucleation requires "seeds" or existing crystals to perpetuate crystal growth. In our salt example, we bypassed primary nucleation by "seeding" the solution with a final teaspoon of salt. Secondary nucleation can be thought of as the workhorse of crystallization.

Again, no complete theory is available to model secondary nucleation and it's behavior can only be anticipated by experimentation. Mathematic relationships do exist to correlate experimental data. However, correlating experimental data to model crystallization is time consuming and often considered extreme for batch operations, but can easily be justified for continuous processes where larger capital expenditures are necessary. For batch operations, only preliminary data measurements are truly necessary.

We've discussed how crystallization occurs once supersaturation is reached, but how do we reach supersaturation? We have already covered one such method in our salt crystallization example. Since the solubility of salt in water decreases with decreasing temperature, as the solution cools, its saturation increases until it reaches supersaturation and crystallization begins (Figure 3). Cooling is one of the four most common methods of achieving supersaturation. It should be noted that cooling will only help reach supersaturation in systems where solubility and temperature are directly related. Although this is nearly always the case, there are exceptions. In Figure 3, you'll note that Ce₂(SO₄)₃ actually becomes less soluble in water at higher temperatures.

Figure 3: Solubilities of Several Solids

The four most common methods of reaching supersaturation in industrial processes are:

1. Cooling (with some exceptions)

2. Solvent Evaporation

3. Drowning

4. Chemical Reaction

In an industrial setting, the solute-solvent mixture is commonly referred to as the "mother liquor".

Drowning describes the addition of a nonsolvent to the solution which decreases the solubility of the solid. A chemical reaction can be used to alter the dissolved solid to decrease its solubility in the solvent, thus working toward supersaturation. Each method of achieving supersaturation has its own benefits. For cooling and evaporative crystallization, supersaturation can be generated near a heat transfer surface and usually at moderate rates. Drowning or reactive crystallization allows for localized, rapid crystallization where the mixing mechanism can exert significant influence on the product characteristics.

Equipment Used in Crystallization

1. Tank Crystallizers

This is probably the oldest and most basic method of crystallization. In fact, the "pot of salt water" is a good example of tank crystallization. Hot, saturated solutions are allowed to cool in open tanks. After crystallization, the mother liquor is drained and the crystals are collected. Controlling nucleation and the size of the crystals is difficult. The crystallization is essentially just "allowed to happen". Heat transfer coils and agitation can be used. Labor costs are high, thus this type of crystallization is typically used only in the fine chemical or pharmaceutical industries where the product value and preservation can justify the high operating costs.

2. Scraped Surface Crystallizers

An example may be the Swenson-Walker crystallizer consisting of a trough about 2 feet wide with a semi-circular bottom. The outside is jacketed with cooling coils and an agitator blade gently passes close to the trough wall removing crystals that grow on the vessel wall.

3. Forced Circulating Liquid Evaporator-Crystallizer

Just as the name implies, these crystallizers combine crystallization and evaporation, thus the driving forces toward supersaturation. The circulating liquid is forced through the tubeside of a steam heater. The heated liquid flows into the vapor space of the crystallization vessel. Here, flash evaporation occurs, reducing the amount of solvent in the solution (increasing solute concentration), thus driving the mother liquor towards supersaturation. The supersaturated liquor flows down through a tube, then up through a fluidized area of crystals and liquor where crystallization takes place via secondary nucleation. Larger product crystals are withdrawn while the liquor is recycled, mixed with the feed, and reheated.

4. Circulating Magma Vacuum Crystallizer

In this type of crystallizer, the crystal/solution mixture (magma) is circulated out of the vessel body. The magma is heated gently and mixed back into the vessel. A vacuum in the vapor space causes boiling at the surface of the liquid. The evaporation causes crystallization and the crystals are drawn off near the bottom of the vessel body.

Compressor Surging Under Control

Turbocompressors, either centrifugal or axial, are the heart of many industrial processes. Often, these compressors are critical to the operation of the plant, yet they are seldom installed with a spare unit. Surging represents a major threat to compressors and these processes. Surge prevention is an important process control problem in these environments as surging can result in costly downtime and mechanical damage to the compressors. An effective anti-surge control system is critical for every turbocompressor.

Understanding Surge

Many believe that surging is analogous to cavitation in a centrifugal pump, but this is not the case. Surging is defined as a self oscillation of the discharge pressure and flow rate, including a flow reversal. Every centrifugal or axial compressor has a characteristic combination of maximum head and minimum flow. Beyond this point, surging will occur. During surging, a flow reversal is often accompanied by a pressure drop.

Surging is best illustrated by observing the movement of the compressor operating point along its characteric curve as shown in Figure 1.

.

Figure 1: Examining the Characteristic Curve

Development of the Surge Cycle

Consider a compressor system as shown in Figure 2. The discharge pressure is marked Pd and the downstream vessel pressure is Pv.

Figure 2: Example Compressor System

Now, referencing Figure 3, assume that the system is operating at steady state at Point D. If the demand for gas is reduced, the operating point will move toward Point A, the surge point. If the load is reduced enough, the compressor operating point will cross Point A. Beyond Point A, the compressor loses the ability to increase the discharge pressure such that Pd will become less than Pv. This is the flow reversal observed during surging. The operating point will then jump to Point B.

Figure 3: Graph of Operating Points

Point B is not a stable operating point. When the flow reversal occurs, the discharge pressure drops. This forces the operating point to move from Point B to Point C. At Point C, the flow rate is insufficient to build the necessary pressure to return to Point A. Thus, the operating point moves to Point D where the flow rate is in excess the load demanded and the pressure builds until Point A is finally reached. This completes a single surge cycle. The next cycle begins again with another flow reversal and the process repeats until an external force breaks the surge cycle.

Consequences of Surging

Consequences of surging can include:

1. Rapid flow and pressure oscillations cause process instabilities

2. Rising temperatures inside the compressor

3. Tripping of the compressor

4. Mechanical damage

Mechanical damage can include:

· Radial bearing load during the initial phase of surging. A side load is placed on the rotor which acts perpendicular to the axis.

· Thrust bearing load due to loading and unloading.

· Seal rubbing

· Stationary and rotating part contact if thrust bearing is overloaded.

Anti-Surge Control

The only way to prevent surging is to recycle or blow down a portion of the flow to keep the compressor away from it's surge limit. Unfortunately, compressing extra flow results in a severe economic penalty. Thus, the control system must be able to accurately determine the compressor's operating point as to provide adequate, but not excessive, recycle flow.

A Surge Limit Line (SLL) is the line connecting the various surge points of a compressor at varying RPMs. The set point of the anti-surge controller is represented on the compressor map shown in Figure 4 by a line which runs parallel to the surge limit line. This line is called the Surge Controller Line (SCL). The controller is then able to calculate the deviation from the operating point to the SCL.

Figure 4: Compressor Operating Map

The compressor surge limit is not fixed with respect to any one measured variable such as compression ratio or pressure drop across the flow meter. Instead, it is a complex function that is dependent on the gas composition, RPM, suction temperature, and pressure. A closed loop PI controller would be unable to prevent surge during large or fast disturbances. Therefore, such a controller would be unable to stop surge. Rather, the controller would simply cycle the recycle valve open and closed in response to successive surge cycles. For a PI controller to act quickly, the "b" value would need to be high. This would result in a decreased operating region for the compressor when the recycle valve is closed.

Thus, an open loop control is used in conjunction with the closed loop in an anti-surge controller. The overall configuration is shown in Figure 5. A Recycle Trip Line (RTL) is used between the SLL and the SCL. Small or slow distrubances are managed by the closed loop controller which keeps the compressor operating point to the right of the RTL. For large or fast disturbances, the compressor operating point will reach the RTL. At this point, the open loop control will be initiated. This will add a step change which is a function of the compressor operating point at the moment it reaches the RTL. In this manner, the fast opening valve will be sufficient to stop surging.

Adaptive gain is also used in the anti-surge controller. When the operating point moves quickly toward the SCL, the adaptive gain move the SCL toward the operating point.

Figure 5: Compressor Anti-Surge Control Scheme

Anti-Surge Valve Requirements

1. The valve must be large enough to prevent surging under all possible operating conditions. However, a valve which is too oversized will result in poor control.

2. Stroke speed - A fast stroke speed is very important.

3. Ensure adequate air supply to properly operate the valve.

4. Tubing run should be minimized to reduce lag time.

5. One or more volume boosters are required to ensure fast response and equal opening and closing time.

6. Fail position should be open

Basics of Industrial Heat Transfer

Heat transfer is one of the most important industrial processes. Throughout any industrial facility, heat must be added, removed, or moved from one process stream to another. Understanding the basics of the heart of this operation is key to any engineers’ mastery of the subject. There are three basic types of heat transfer: conduction, convection, and radiation. The two most common forms encountered in the chemical processing industry are conduction and convection. This course will focus on these key types of heat transfer.

Any overall energy balance starts with the following equations:

Where:

Q = heat transferred in thermal unit per time (Btu/h or kW)

M = mass flow rate

T = temperature

Cp = heat capacity or specific heat of fluid

Subscript “H” = hot fluid

Subscript “C” = cold fluid

In theory, the heat given up by the hot fluid is never exactly equal to the heat gained by the cold fluid due to environmental heat losses. In practice, however, they are generally assumed to be equal to simplify the calculations involved. Any environmental losses are generally minimized with insulation of equipment and piping.

When examining industrial systems, it is common practice to use a graphical form of these equations know as “T-Q diagrams” to enhance understanding and to make sure that the Second Law of Thermodynamics is not disobeyed. In other words, heat can only move from a higher to a lower temperature fluid. Here is how the generic diagram is constructed:

It’s easy to see how viewing a particular heat transfer problem in this way is extremely valuable.

Now that’s we’ve seen how heat moves from a hot fluid to a cold fluid, let’s examine the third basic equation that is used to govern the equipment used for transferring heat.

The “Heat Exchanger Equation” takes the form:

Where:

Q = heat transferred in thermal unit per time (Btu/h)

f = temperature correction factor

U = overall heat transfer coefficient (Btu/h ft² °F)

A = heat transfer area (ft²)

LMTD = log mean temperature difference

These three (3) equations are the basis for virtually all heat exchanger design.

Examining the “Heat Exchanger Equation”

If we take a closer look at the heat exchanger equation, it’s worth noting some assumptions that are made in its derivation. First, the overall heat transfer coefficient and the specific heat (also called heat capacity) of the fluids are assumed to remain constant through the heat exchanger.

If we look at the change in the heat capacity of water, for example, over a reasonable tempe

rature range, here is what we find:

Specific heat of water at 100 °F and atmospheric pressure = 0.9979 Btu / lb °F

Specific heat of water at 210 °F and atmospheric pressure = 1.0066 Btu / lb °F

So, we can see that this is a fairly reasonable assumption for water and it remains reasonable for most industrial fluids. The specific heat of a substance is defined as the amount of heat required to raise the temperature of one pound of the substance by a single degree Fahrenheit (other units can apply as well).

The overall heat transfer coefficient is a calculated variable based on the physical properties of the fluids involved in the heat transfer (hot and cold) as well as the geometry and type of heat exchanger to be used. We’ll examine this c

loser a little later.

The log mean temperature difference or LMTD is used to describe the av

erage temperature difference throughout the exchanger. The difference between the temperatures of the fluids provides the “driving force” for the heat transfer to occur. The larger the temperature difference, the smaller the required heat exchanger and vice versa.

You’ll notice from our T-Q diagram used to explain the equations:

that it appears that the temperature difference between the fluids remains almost constant throughout the heat exchanger. This is rarely the case. Let’s look at a more practical example. Let’s assume that a process stream containing water at 200 °F is to be cooled to 150 °F using cooling tower water available at 85 °F. It is common practice in industry to return cooling tower no higher than 120 °F. In other words, the cooling tower water flow must be such that its outlet temperature from the heat exchanger is less than 120 °F. The reason for this is that cooling tower water often contains treatment chemicals that can plate out onto heat transfer surfaces and cause severe fouling or degradation of the heat transfer rate at elevated temperatures.

Here is what the T-Q diagram may look like for our example case:

You can see that the temperature difference between the two streams will vary widely. This is why the log mean temperature difference is used. Here is how the log mean temperature difference works

So, for a heat exchanger as described above, we calculate the LMTD as follows:

There can be special cases where the LMTD equation shown above is not applicable. Consider the case below.

If you tried apply the LMTD equation to this special case, you’d find that the result would be zero. In this case the LMTD is the same as the temperature difference on each “end” of the heat exchanger, or 100 °F.

A Brief Word on Flow Direction

Notice that up to this point, the two fluids considered in a heat exchanger have been moving in opposite directions to one another. This is known as counter-current flow. This is the predominantly preferred flow direction because it results in higher temperature difference driving forces within the heat exchanger, thus minimizing the heat transfer area required.

The other flow configuration, where the fluids flow in the same direction, is called co-current flow. Co-current flow, while it is rarely used, does have the advantage of lowering the heat exchanger wall temperature on the hot side fluid. This can be useful for temperature sensitive fluids or as a means of minimizing deposits that are temperature sensitive.

The Temperature Correction Factor, f

The temperature correction factor, f, is used to correct the log mean temperature difference for heat exchangers than lack truly counter-current flow. Many different heat transfer technologies lack truly counter-current flow patterns as a result of their inherent mechanical design. Generally, the value for f should be between 0.75 of 0.97. There are cases when this value can be taken as one, but only if the flow in the exchanger is purely counter-current. There are countless charts available to look up the temperature correction factor for a given configuration.

The Overall Heat Transfer Coefficient

The overall heat transfer coefficient describes the rate of heat transfer in the heat exchanger. Generically, it is described by the following equation:

Where:

U = overall heat transfer coefficient (Btu / h ft² °F)

h_H = hot side heat transfer coefficient

h_C = cold side heat transfer coefficient

Delta x = exchanger wall thickness

k = exchanger wall material thermal conductivity

R_f = fouling coefficient (h ft² °F / Btu)

The equation for the overall heat transfer coefficient is often reduced to the following:

because the term Delta x / k seldom has any significant impact on the overall U-value.

The overall heat transfer coefficient can either be calculated, looked up in reference materials for a given duty, estimated from past plant experience, or supplied by a heat exchanger vendor.

Brief Overview of Heat Exchanger Types

In the chemical processing industry, there are numerous types of heat exchanger devices. The types of exchangers can be classified by the duty that they perform, surface compactness, construction features, flow arrangements, and others. In general, a heat exchanger can fall into one of these processing categories:

No Phase Change

Liquid to Liquid heat transfer

Liquid to Gas heat transfer

Gas to Gas heat transfer

Phase Change

Condensing a vapor with a liquid or gas service fluid

Vaporizing a liquid with a liquid, gas, or condensing fluid

Heat exchangers can also be broken down into the following two types of mechanical geometries:

Shell and Tube Heat Exchangers

Compact and Extended Surface Heat Exchangers

Approximately 70-80% of the heat exchanger market is dominated by the shell and tube type heat exchanger. It is largely favored due to its long performance history, relative simplicity, and its wide temperature and pressure design ranges. We will explore this technology in further detail later.

The second category mentioned, compact and extended surface heat exchangers, play a smaller role in the chemical processing industry. Some of the available technologies that fit into this category are the plate and frame heat exchanger, finned tube heat exchangers, spiral heat exchangers, fin-fan heat exchangers, and many others.

Compact Heat Exchanger Technologies

The plate exchanger, shown below, consists of corrugated plates assembled into a frame. The hot fluid flows in one direction in alternating channels while the cold fluid flows in true countercurrent flow in the opposite alternating channels. The fluids are directed into their proper channels either by a rubber gasket or a weld depending on the type of exchanger chosen.

Traditionally, plate and frame exchangers have been used almost exclusively for liquid to liquid heat transfer. Today, many variations of the plate technology have proven useful in applications where a phase change occurs as well. This includes condensing duties as well as vaporization duties. Plate heat exchangers are best known for having overall heat transfer coefficients (U-values) in excess of 3-5 times the U-value in a shell and tube designed for the same service.

Plate exchangers can be especially attractive when more expensive materials of construction are required. The significantly higher U-value results in far less area for a given application, thus a lower purchased and installed cost due to its relatively small size. The higher U-values are gained by inducing extremely high wall shear on the plate surface. The best way to think of a plate heat exchanger is that it is essentially a static mixer that happens to transfer heat very well. The plate exchanger, by virtue of its high wall shear stress also minimizes fouling very well.

Typical plate thicknesses range from 0.40 mm to 0.60 mm and passage channel openings can range from 1.5 mm up to 11.0 mm depending on the application and required design pressure (the larger the opening, the lower the design pressure available). These small passages also restrict the size of solids that can be successfully passed through the exchanger.

Perhaps the biggest advantage of the plate and frame heat exchanger, and a situation where it is most often used, is when the heat transfer application calls for the cold side fluid to exit the exchanger at a temperature significantly higher than the hot side fluid exit temperature. This situation is best explained with another set of T-Q diagrams:

Duty 1 shown above is easily accomplished in a single and tube heat exchanger.

Duty 2 shows a severe “temperature cross” or the cold side fluid exiting higher than the hot side fluid. This would require several shell and tube exchangers in series due to the lack of purely counter-current flow. On the other hand, this duty is easily accomplished in a single plate and frame heat exchanger.

Finned tube heat exchangers are commonly used to transfer heat between a gas and liquid. The tubes used in these units are equipped with fins that extend outward from the tubes as shown below.

The fins on the tubes allow for a much larger surface area to be packed into a small volume. This is especially important when transferring heat to or from a gas as gasses have extremely low heat transfer coefficients (meaning that large amounts of area are required).

Fin-fan heat exchangers are designed to use air to cool process fluids. Think of them as a giant radiator. The process fluid is passed through the coils and a fan helps pull air over the outside surface to promote cooling. These units again must provide a very large surface area to make up for the poor heat transfer of the air.

Shell and Tube Heat Exchanger Technologies

Shell and tube heat exchangers are known as the work-horse of the chemical process industry when it comes to transferring heat. These devices are available in a wide range of configurations as defined by the Tubular Exchanger Manufacturers Association. In essence, a shell and tube exchanger is a pressure vessel with many tubes inside of it. One process fluids flows through the tubes of the exchanger while the other flows outside of the tubes within the shell. The tube side and shell side fluids are separated by a tube sheet.

The shell and tube type is usually indicated as a three (3) letter code from the TEMA specifications shown below:

The shell side of a shell and tube exchanger usually contains baffles as shown above to direct the shell side flow around the tubes to enhance heat transfer. As you can see, shell and tube exchangers can be configured for liquid-liquid, gas-liquid, condensing, or vaporizing heat transfer.

The tubes can be a different material than shell and the shell can either be cladded or of solid construction. It’s impossible to go over all of the mechanical details of the shell and tube here, but this should provide you with a general overview of the construction. There are numerous other sources of information freely available on these types of units.

The tubes and shell can be designed for a variety of design temperatures and pressures.

The thermal design of shell and tube heat exchangers is often performed by vendors. The process engineer generally completes a TEMA specification sheet and submits it to vendors for bids. If you’re interested in more details on the thermal design aspects of shell and tube heat exchangers

There are well documented sources of estimated overall heat transfer coefficients and fouling factors that can be specified. Fouling factors are historic safety factors that allow for the oversizing of a shell and tube in anticipation of eventual surface build-up that will form a resistance to heat transfer. Remember, the overall heat transfer coefficient of a new heat exchanger will slowly degrade over time until it “levels off” to what is known as the “service U-value”. This is the actual rate of a heat transfer that the unit will achieve on a nominal basis. The combination of a well selected U-value and a fouling factor should ensure a good shell and tube design. Typical U-values for various services and fouling factors can be found on the internet or in various text references.

Included with this course is a MS Excel spreadsheet that can be downloaded and used to specify heat transfer equipment. There is a version which uses U.S. Customary units as well as a version which employs SI units.

Understanding the basics of industrial heat transfer will help you better understand opportunities for cost savings in your plant. With energy prices showing no sign of declining, a good basis in heat transfer will help you calculate just how much you can save by installing a new heat exchanger in your plant. With the use a T-Q diagram and a basic understanding of the equipment available to you, making the right choice in heat transfer equipment can yield results for years to come.