| Home | | | Scope | | | Types/Configs. | | | Schematics | | | Heat Balance | | | Heat Trans1 | | | Heat Trans2 | | | Process | | | Circulation | | | Other | | | Publications | | | Calculators |
| D-Frame evaporator layout |
| O-Frame evaporator layout |
| A-Frame evaporator layout |
| I-Frame evaporator layout |
| Horizontal tube evaporator layout |
| Superheater configurations |
| Economizer configurations |
 |
D-Frame evaporator layout. This configuration is very popular for HRSG units recovering heat from small gas turbines and diesel engines. It is a very compact design and can be shipped totally assembled. It is limited, however, since the bent tube arrangement quickly causes the module to exceed shipping limitations for units having a large gas flow. |
 |
O-Frame evaporator layout. This configuration has probably been used for more years than any of the others. It has the advantage of the upper header being configured as the steam separation drum. Or, the upper header can be connected to the steam drum by risers, allowing more than one O-Frame evaporator to be connected to the same steam drum, resulting in shipable modules being able to handle very large gas flows. |
 |
A-Frame evaporator layout. This configuration is simply a variation of the O-Frame Evaporator. It was popular for services with a large amount of ash, since the center area between the lower drums could be configured as a hopper to collect and remove solid particles. |
 |
I-Frame evaporator layout. In the past twenty years, this configuration has become the most popular of all the Evaporator designs. This type module can be built in multiple axial modules or in multiple lateral modules, allowing it to be designed to accept any gas flow. There are numerous variations of this design where tube bundles may contain one, two, or three rows of tubes per header. It is also, normally, more economical to manufacture, ship and field construct. The tube bundles may be shipped to field installed in the modules, or as loose bundles which are installed into a field erected shell. |
 |
Horizontal tube evaporator layout. The horizontal tube evaporator is used, not only for heat recovery from Gas Turbine exhaust, but for recovery from flue gases in Refinery and Petrochemical furnaces also. It has similar size limitations due to shipping restrictions similar to the O-frame modules. It is generally a less expensive unit to manufacture than the other configurations, but if it is a natural circulation design with large tubes, such as in some CO Boilers, or very long tubes, special consideration needs to be given to assure all tubes are provided with sufficient effluent. These considerations will be discussed later on in this document. |
Superheater designs would normally follow along with the evaporator type that is being used. Three basic superheater designs are shown below, Horizontal Tube, Vertical Tube, and I-Frame. The Horizontal Tube design would normally be used for the D-Frame Evaporator if gas flow is vertical up at the outlet. This horizontal design would be expected to be used also on a horizontal evaporator design. The Vertical Tube design would generally be used with the A-Frame or O-Frame Evaporator and with the D-Frame if the gas exits horizontally. The I-Frame Superheater would be used with the I-Frame Evaporator, but may also be used with the other evaporator designs.
 |
 |
 |
Economizer designs would normally follow along with the evaporator type that is being used and be similar in design to the superheater. The configurations would be similar to the ones shown above for the superheaters.
After deciding on the evaporator type to be used for the unit, the next important step in the design of an HRSG unit is to decide the arrangement of the various coils in the unit. Of course, if only an evaporator is present, this may consist of a very simple schematic, but if, as in most cases, there are more than one coil, then consideration needs to be given as to their position in the gas stream.
Arrangement of coils.Obviously, the best place to put the highest temperature coil, the superheater, would be in the hottest part of the gas stream. Since, this is where it would take the least amount of surface to exchange the heat, and would allow a stepped heat recovery for maximum heat exchange. The curve below shows this relationship between the heat given up and the three primary coils found in an HRSG.
 |
In viewing this generalized sketch showing the relationship between the heat
absorbed and the heat given up, it is easy to see the area referred to as the
"pinch" at the evaporator outlet. By laying a straight edge on the heat given up line and rotating it while holding it at the pinch, it is also, easy to see that, at a very high inlet temperature, there may be a critical approach temperature occur at the economizer inlet, and going the other way, at a lower inlet temperature, this may occur at the superheater outlet. |
Of course, modern HRSG units are not always this simple. The components can and are placed in many configurations to achieve desired results. The range of arrangements that the coils may be placed, is only limited by the users imagination and the constraints of the temperature approaches. Shown below are just a few examples of various arrangements.
Typical single pressure arrangements. |
 |
 |
| Typical dual pressure arrangements. |
 |
 |
 |
 |
 |
Now that we have a general idea of how to arrange the coils, we prepare the flow schematic. This flow schematic gives us a preliminary picture of how the HRSG will look. Also, we can use the sketch to perform the preliminary heat balance which we will review in Section 4.
For our example flow schematic, we will assume a single pressure HRSG with a superheater and economizer section.
 |
It is not important that you necessarily use this style schematic, but it is important to be consistent in the style you use. If you always present an evaporator in the same way, and a superheater always looks like "your" superheater, the flow schematics become very recognizable to anyone needing to refer to them. Remember, the flow schematic does not need to represent the actual mechanical design of the HRSG, neither in looks, or direction of flow, hot to cold, etc.
Now, using a similar approach to above, let's construct a flow schematic to represent a triple pressure unit with an integral deaerator.
 |
The evaporator pinch, or approach temperature, is what limits the amount of heat that can be recovered in most HRSG designs. As was discussed in the previous section, Schematics, the limiting effect of this approach is important. For many general purpose HRSG's such as those found in refineries and chemical plants, a pinch of 50 °F provides an economical design with a realistic payout. But in the more competetive markets of combined cycle or co-generation plants, it is not uncommon to see pinch points below 30 °F. And as a practice, a 30 °F pinch design for these HRSG's should be considered.
It should, however, be remembered that the closer the pinch, or approach, the less reliable the results will be. In other words, it would be easy to calculate the steam generated in a unit at a 5 °F pinch, but the probability of achieving this result with the actual equipment would be almost nil. If you look at the added amount of surface required to go from a 10 °F to a 5 °F pinch versus the change in surface to go from 50 °F to 45 °F, you will quickly see the why this is true.
Other process approach temperatures:Other process appoach temperatures are similar to the special case of the "pinch" discussed above. But, they do not, except in some situations, control the overall design of the HRSG. A 50 °F approach is a good minimum to consider for coils such as hot oil, superheaters, economizers, etc. Of course, the same is true with these coils, the higher the approach temperature, the less surface it will take to exchange the heat. This is why most of the flow in these coils are counter current to the gas flow, which provides a higher approach temperature.
Economizer water approach:The economizer water approach temperature to the evaporator satuaration temperature is very important and should be selected with care. If too close an approach is used in the design, vaporization may occur in the coil during off design cases which may cause severe upsets in the unit. It should be noted, however, that just because the economizer does vaporize at some operating condition, it does not necessarily mean a problem, since the design can be such that it can handle this condition. But, for most designs, it is better to avoid this condition. A normal design approach temperature is 20 °F. This approach gives significant safety factor for load swings. But, again, you should rate HRSG at all expected operating conditions.
Superheated Steam Desuperheating:Superheat desuperheating is the best way to control the outlet temperature of the HRSG superheater. It is not, though, the only way. Steam bypassing around all or part of the superheating coil and then remixing it to control the temperature is done with great success. If a spray desuperheater is used, it can be placed at the outlet, or at an itermediate point in the superheater coil. Placing it at an intermediate point gives the added protection of preventing accidental water slugs which may damage downstream equipment.
Blowdown requirements:The boiler blowdown requirement is set by the condition of the feedwater. Primarily it is used to control solids build up in the steam separation drum. If nothing is known of the feedwater at time HRSG is being designed, an allowance should be used in design. For normal modern facilities, a 2% allowance should be sufficient. For others, a 5% allowance should be provided for in the design. But, you should keep in mind that somewhere along the route from design to production, this must be revisited to assure proper operating conditions in the HRSG.
We begin with the first sample schematic that we prepared in Section 3, a single pressure HRSG with a superheater, evaporator, and economizer.
|
For our process conditions, we will assume the following:
Gas Side :
800,000 lbs/hr of Gas Turbine Exhaust at 980 °F
Setting Loss To Atmosphere,
2% of Heat Absorbed
Maximum Back Pressure at Gas Turbine Exhaust Flange, 8"
H2O
| Gas Properties : Volume % | |
| Nitrogen, N2 | 72.55 |
| Oxygen, O2 | 12.34 |
| Carbon Dioxide, CO2 | 3.72 |
| Water, H2O | 10.52 |
| Argon, Ar | 0.87 |
| Sulphur Dioxide, SO2 | 0.0 |
| Carbon Monoxide, CO | 0.0 |
Tube Side : Steam at outlet, Maximum Flow at 600 psig and 750 °F
Feedwater
at 227 °F and pressure required at inlet.
| Pinch At Evaporator, °F | 50.0 |
| Economizer Water Approach, °F | 20.0 |
| Blowdown, % of Steam Out | 2.0 |
| Pressure Drop In Superheater, psi | 15.0 |
| Pressure Drop In Economizer, psi | 10.0 |
Now, we have set all of our conditions, so we can proceed with a heat balance. For these calculations, we will need a calculator to provide us with the properties of the flue gas, water, and steam. We can start these in separate windows so we can keep them available as we work out our solution.
We can now populate our schematic with all known values.
 |
Now we can calculate the missing data.
Heat Available To Evaporator And
Superheater:
We can now complete our schematic with all known values.
 |
Selecting the tube material and size to use in a HRSG design is really a matter of experience. As you work with different HRSG's for different services, you develop a knowledge of what fit before in a similar design, so you know where to start with a new design. But a few general rules can be used to start the selection.
For the typical, general purpose HRSG, using standard tubing sizes, the 2" tube size will normally work out to be the most economical tube size. The cost will generally go up with a smaller or larger tube size. Most HRSG units recover heat from a relatively low temperature gas, i.e., less than 1,000 °F. Of course, many of the modern HRSG's are supplementary fired to achieve even greater efficiencies. But, with the exception of the superheater, you can normally assume that carbon steel tubes will work for the evaporator and the economizer. If the superheater outlet temperatures are low, such as 600 °F and below, you should be able to assume carbon steel tubes to start. If higher than 600 °F, you may want to start with T11 tubes.
In a similar manner, you can make some preliminary estimates to determine what the design metal temperature for the HRSG tubes need to be. With this temperature, you would select the least material that is good for the temperature. Eventual analysis may show that a higher alloy and a thinner wall may be more economical, so running calculations with several materials is always wise.
Typical generic, pipe and tube specifications used for HRSG tubes:| Generic Specification | Pipe Specification | Tube Specification |
| Carbon Steel | SA 106 Gr B | SA 178 A |
| 1¼ Cr ½ Mo | SA 335 Gr P11 | SA 213 T11 |
| 2¼ Cr 1 Mo | SA 335 Gr P22 | SA 213 T22 |
| 5 Cr ½ Mo | SA 335 Gr P5 | SA 213 T5 |
| 9 Cr 1 Mo | SA 335 Gr P9 | SA 213 T9 |
| 18 Cr 8 Ni | SA 312 TP 304 | SA 213 TP 304 |
| 16 Cr 12 Ni 2 Mo | SA 312 TP 316 | SA 213 TP 316 |
| 18 Cr 10 Ni Ti | SA 312 TP 321 | SA 213 TP 321 |
| 18 Cr 10 Ni Ti | SA 312 TP 321H | SA 213 TP 321H |
And other, more exotic materials for special services are used as may be needed. The wall thickness required, for the heat absorbing tubes, is calculated by using the ASME Section 1. For heat absorbing tubes, there are two formulas that may apply, so it is normal practice to check the required thickness and maximum allowed working pressure, MAWP, using both formulas, then using the more appropriate.
Tube Wall Thickness:| t = Minimum required thickness, in |
| P = Maximum allowable working pressure, psia |
| D = Outside diameter of cylinder, in |
| S1 = Maximum allowable stress value (PG-23), psi |
| S2 = Maximum allowable stress value (PG-23), psi |
| e = Thickness factor for expanded tube ends |
| y = Temperature coefficient |
| E = efficiency |
| C = Minumum allowance for threading and structural stability, in |
Using the above stress values and formulas, we can now calculate the minimum wall thickness for a tube.
In the above tube wall thickness calculator, only the "Tube Specifications", such as SA-178 Gr. A, are being used for stress values, so this calculator is not valid if using a "Pipe Specification", such as SA-106 Gr. B, since the stress values are different. A little bit of confusion over why a pipe is called a tube, when it is in a HRSG, might be expected. But usage dictates that the heat transfer tubes be referred to as tubes regardless of whether they are manufactured from materials specified as tubes or as pipes. If the use is as a downcomer or riser, etc., it is called a "pipe".
When using tubes with the OD of a standard pipe size or using piping specifications and the HRSG uses returns, you would normally select a standard return bend to return the flow to the next tube. These returns bends are normally manufactured in two turning radii, called "short radius" and "long radius". The short radius return refers to a 180° return bend using a radius of one nominal diameter, ie, a 4" pipe size return has a radius of 4", and a 6" has a radius of 6", etc. The "long" radius bend has a radius equal to 1.5 nominal diameters so a 4" return has a radius of 6" and a 6" has a radius of 9". These standard returns are manufactured in most of the pipe schedules and are also available in "minimum wall" specifications.
Typical generic and pipe specifications used for return bends:| Generic Specification | Pipe Specification |
| Carbon Steel | SA 234 WPB |
| 1¼ Cr ½ Mo | SA 234 WP11 |
| 2¼ Cr 1 Mo | SA 234 WP22 |
| 5 Cr ½ Mo | SA 234 WP5 |
| 9 Cr 1 Mo | SA 234 WP9 |
| 18 Cr 8 Ni | SA 403 WP304 |
| 16 Cr 12 Ni 2 Mo | SA 403 WP316 |
| 18 Cr 10 Ni Ti | SA 403 WP321 |
| 18 Cr 10 Ni Ti | SA 403 WP321H |
If using standard pipe fittings manufactured to standard pipe schedules, you would assume 80% of the standard wall thickness. If you are bending tubes or pipe for the application, you would need to calculate the thinning in the bend. The following calculator estimates the ratio of that thinning.
Pipe-Wall Thinning In Bends:Now that we have selected a tube diameter, material, wall thickness, and tube spacing, we need to decide what length the tubes should be. Pipe and tubes are manufactured in random lengths, ie, since the billet size varies, the actual length of the tube that is extruded, from a billet, varies from one tube to the next. For lower cost materials, it is usually cheaper to scrap pieces of tube, then it is to make center welds to try and use all the material. But another high cost factor involved with the length is the supports and guides for the tube in the HRSG.
For vertical tubes, usually the overall HRSG shape and size dictate the best tube length. It is necessary to consider the maximaum shipping width and length in setting the tube length. The support and guide requirement varies depending on whether the tubes are supported from the top(hung) or bottom of the tubes.
In the horizontal tube HRSG's, the overall shape and size also figure into the equation. But, within these constraints, the span between supports must be considered. If the user has not specified a maximum span, then generally you would not want to exceed 35 tube OD's. This has been a general industry "not to exceed" rule of thumb used for many HRSG designs. But care should be taken to consider the service and wall temperature of the tubes. Once you have determined the span between supports, the tube length would be selected to use the minimum number of supports, while avoiding unnecessary centerwelds, if centerwelds are allowed by user. All of this must be balanced with the fact that the pressure loss in the tubes is increased dramatically in the returns, so generally you want the longest straight tube possible. The pressure loss in the returns is reviewed in the "Process" section , under "Intube Pressure Drop".
Now, using the single pressure HRSG that we demonstrated in developing the heat balance in Section 4, we can select the heat transfer tubes necessary to proceed with the thermal design. We are going to select an HRSG unit using 2.000" od tubes, so the following selections will be based on this tube od.
|
For the superheater, we can assume that the tube wall temperature will be above 750 °F and less than 850 °F, so we will use the 850°F for the design temperature. We will select for our HRSG design, the O-Frame evaporator. So for the superheater we could use a Vertical Tube or an I-Frame type. We will choose the I-Frame.
|
This design can use either bent tubes, i.e., three rows per bundle, or two straight tubes in a bundle. For this sample, we decide we want three rows per bundle using an 8" pipe header, then we we will decide that the bend radius of the bend in the first and third rows is 6". Using a design pressure of 700 psia, we can check to see if a standard 0.120" minimum wall thickness will be sufficient for this design. Using our bend thinning calculator, we see that the wall thickness after bending will be 0.1013". Then using the tube wall thickness calculator, we see that using the PG 27.2.1 method, that with our 1/32" corrosion allowance, the tube thickness is okay. |
Now, we look at our evaporator, the O-Frame design that we selected above.
|
Okay, here again we have bent tubes. We will assume that the bend radius of these tubes is 12", so the wall thickness using 0.120" to start with, will be 0.1091" after bending. Using SA 178 Gr A tubes, we check the required wall thickness and find that this tube will be fine. Note that setting the design temperature below 700° F really doesn't have any affect on the PG 27.3.1 calculation, since this formula requires a minimum of 700 °F for the stress value selection. When we check the required wall, we get 0.1102 if we use a 1/32" corrosion allowance. but since no corrosion allowance was indicated, we will use this tube. |
For our economizer, we will use the I-Frame design, but use two tubes per bundle with 6" pipe headers, so we will once again need bent tubes, using a 6" bend radius. This results in a 0.1013" wall thickness at the bends. For this economizer, the 0.120' wall thickness will be okay without corrosion allowance.
as been a general industry "not to exceed" rule of thumb used for many HRSG designs. But care should be taken to consider the service and wall temperature of the tubes. Once you have determined the span between supports, the tube length would be selected to use the minimum number of supports, while avoiding unnecessary centerwelds, if centerwelds are allowed by user. All of this must be balanced with the fact that the pressure loss in the tubes is increased dramatically in the returns, so generally you want the longest straight tube possible. The pressure loss in the returns is reviewed in the "Process" section , under "Intube Pressure Drop".Now, using the single pressure HRSG that we demonstrated in developing the heat balance in Section 4, we can select the heat transfer tubes necessary to proceed with the thermal design. We are going to select an HRSG unit using 2.000" od tubes, so the following selections will be based on this tube od.
|
For the superheater, we can assume that the tube wall temperature will be above 750 °F and less than 850 °F, so we will use the 850°F for the design temperature. We will select for our HRSG design, the O-Frame evaporator. So for the superheater we could use a Vertical Tube or an I-Frame type. We will choose the I-Frame.
|
This design can use either bent tubes, i.e., three rows per bundle, or two straight tubes in a bundle. For this sample, we decide we want three rows per bundle using an 8" pipe header, then we we will decide that the bend radius of the bend in the first and third rows is 6". Using a design pressure of 700 psia, we can check to see if a standard 0.120" minimum wall thickness will be sufficient for this design. Using our bend thinning calculator, we see that the wall thickness after bending will be 0.1013". Then using the tube wall thickness calculator, we see that using the PG 27.2.1 method, that with our 1/32" corrosion allowance, the tube thickness is okay. |
Now, we look at our evaporator, the O-Frame design that we selected above.
|
Okay, here again we have bent tubes. We will assume that the bend radius of these tubes is 12", so the wall thickness using 0.120" to start with, will be 0.1091" after bending. Using SA 178 Gr A tubes, we check the required wall thickness and find that this tube will be fine. Note that setting the design temperature below 700° F really doesn't have any affect on the PG 27.3.1 calculation, since this formula requires a minimum of 700 °F for the stress value selection. When we check the required wall, we get 0.1102 if we use a 1/32" corrosion allowance. but since no corrosion allowance was indicated, we will use this tube. |
For our economizer, we will use the I-Frame design, but use two tubes per bundle with 6" pipe headers, so we will once again need bent tubes, using a 6" bend radius. This results in a 0.1013" wall thickness at the bends. For this economizer, the 0.120' wall thickness will be okay without corrosion allowance.
The heat transfer sections of the HRSG frequently use extended surface to improve the overall heat exchange between the hot gases and the steam or water in the tubes. These extended surfaces are usually either a thin plate fin wraped helically around the tube or round or eliptical shaped studs. Following is a description of the more popular extended surfaces.
Segmented Fins:| High Frequency Continuously Welded |
 |
 |
| Standard Frequency Spot Welded |
 |
 |
The standard frequency, spot welded, design is not used as often since this design is normally selected when using very thin, high density finning such as in a large heat recovery boiler. Most HRSG designs use fins less than 0.049 inch thick. The standard frequency, spot welded fin also has a foot which presents a place where corrosion can occur if flue gases are corrosive or moisture is present.
Solid Fins:| High Frequency Continuously Welded |
 |
 |
| Resistance Welded |
 |
 |
Thermal rating procedures for all these extended surface types are presented in the following pages. Both segmented fin types are rated using the same formulas.
For our sample boiler, that we have been developing throughout this material, we will choose to use 0.049" thick fins with a density of 6 fins per inch. We will use an 11% chrome alloy material for the superheater, A-176 TP409. For the evaporator and the economizer, we will use carbon steel, A 366, which should be good for all the temperatures in these sections. We will recheck our selections after doing the thermal calculations to confirm they are okay. It is important to note, that since the source of our gas is a gas turbine, it is a very clean service and we could have used a thinner fin at a higher density.