Author: Bruce Wernick PrEng., BScEng.

Last Updated: 10 March 2010

I have been trying to automate the coil circuiting process for some time now and come to the conclusion that this is an art rather than a science. The purpose of this article is to document the rules for correct circuiting and to show some examples of correctly laced coils.

I am referring particularly to the heat exchanger most commonly used in the air conditioning industry. This is a fin-and-tube heat exchanger commonly called the coil, clearly because a tube is coiled into a compact arrangement.

The fluid on the outside is air and due to the relatively low heat transfer coefficient, needs to be enhanced. This is done by extending the surface area through the use of fins. Each of the tubes could be individually finned as shown below, but this turns out to be an expensive option.

figure 1. Finned tube

A more usual configuration is to punch tube holes in a thin aluminium plate and lace the tubes through these holes to form a core block.

figure 2. Coil core block

The question now is how to link the tubes and this is the subject that we deal with in this paper.

One simple case would be link every tube on one side of the coil to a common inlet header and to link the other sides to a common return header.

figure 3. Trivial case of a single tube pass

So, why is this trivial case not ideal?

For practical reasons, we would prefer to have the headers on the same side of the coil. In real installations, you don't always have easy access to both sides of the coil and you would need to be able to get to the connections during servicing.

An equally important reason is that by having a once through system, the tube
side fluid is divided equally across all of the tubes. This means that you
would have the lowest possible pressure drop but it also means that you would
have the __lowest heat transfer__. This is because the velocity of
the tube side fluid is a minimum.

Well, how about the other extreme? Why not connect all the tubes in series since this would give us the highest heat transfer? Yes, it would be possible to make a single coil but it would unfortunately also result in the highest pressure drop.

figure 4. Coil with single circuit

From this you should realize that the actual coil circuiting is a compromise between tube side pressure drop and heat transfer.

As a rule of thumb, you could define arbitrary pressure drop limits and blindly calculate the number of parallel circuits such that you achieve these values. In principle, this is what we do and table 1 gives some guidelines for this process.

Coil Type | Value | Units |

Chilled Water Coil | 25 | kPa |

Evaporator Coil | 40 | kPa |

Condenser Coil | 80 | kPa |

There are however some complications that you should be aware of.

You should have the same number of tubes in each parallel circuit. Failure to achieve this would result in an uneven flow distribution through the coil. Now think of a our previous proposal to achieve the limiting pressure. From the design duty, we would know the required fluid flow. A coil performance calculation would result in a certain number of tubes in the coil. In theory, we can divide the number of tubes into circuits such that the design pressure drop is achieved. The problem is, there is an fixed number of tubes and they generally don't always divide out evenly.

In addition, to ensure that we get the headers on the same side of the coil, the number of tubes in a circuit must be an even number.

Complicated enough for you yet?

As we consider the different types of coils, we find that there are other considerations that must be kept in mind.

Some rules that I have used are as follows:

You should not have traps in any of the circuits. In other words, it must allow the tube side fluid to drain from the coil by gravity.

Chilled water and hot water coils are circuited upwards from inlet to outlet. This is to ensure that air locks do not form. At the water coil outlet header, you also have a bleed to remove the air.

Direct expansion coils are circuited upwards from inlet to outlet. This means that as the entering liquid evaporates, the vapor forms at the top outlet end.

Air cooled condenser coils are circuited downwards from inlet to outlet. This is to ensure that condensed liquid refrigerant can drain freely from the bottom of the condenser and that the liquid can not run back to the compressor valves when the system stops.

The circuit count is based on tube side pressure drop. A single circuit will result in the maximum tube side pressure drop. The idea is therefore to select the number of circuits such that the reasonable pressure drops occur.

In the DX coil, there is an additional requirement. The velocity must be high enough to carry oil back to the compressor. This means that there is a minimum number of circuits on the DX coil.

For logical circuiting we define a serpentine number. This is measured by circuits/tubes high. For example, the figure alongside shows a 4-row coil with 12 tubes high and 6 circuits. The coil serpentine is 6 circuits/12 tubes = 1/2s |

Typical serpentine values are 1/4s, 1/3s, 1/2s, 5/8s, 2/3s, 3/4s, 5/6s, 1s, 4/3s, 3/2s, 2s.

By using return bends at each end, the tubes are laced into a circuit such that the rules above are applied. The best way to get to grips with circuiting is to look at actual circuit drawings.

You can only link adjacent tubes. The reason is that the return bends only come in two sizes. Ideally, you should try to get as many hairpin bends as possible since this reduces the number of welds needed per coil.

figure 5. Tube pattern

Finally you can't cross circuits. This would just make a real mess of your header connections.

The table below show a few examples of coil circuits that follow the above rules. These are not unique, you could lace these coils in many ways to achieve the same objectives and this is the reason that it is so difficult to automate the task.

| |||

half | three quarter | single | |

2 row | |||

3 row | |||

4 row | |||

5 row | |||

6 row |

The concept of serpentine is to identify minimum typical repeating patterns if tube lacings. But I have discovered that there is a much more practical way to categorize coils. That is by using the tubes per circuit. Actually, it comes to the same thing as serpentine but it is much easier to identify (simple count the tubes in a circuit).

Let's look at the numbers.

The coil has **nth** tubes high, **r** rows deep and **c** parallel
circuits.

This means that there are a total of **nt = nth x r** tubes in the coil.

By definition, serpentine **s = c / nth**

Clearly, we can also define the number of tubes per circuit **tpc = nt / c**

and now we can relate **s** to **tpc** by substituting for **nt**
and **c** in the **tpc** equation.

tpc = nt / c = (nth x r) / (nth x s) = r / s

So, re-arranging for s shows that s can be expressed in terms of tpc

**s = r / tpc**

The difference is, s is a ratio whereas tpc is an integer count.

Check is out with the 4 row coil

r = 4

nth = 8

c = 4

s = c / nth = 4 / 8 = 1/2 serpentine

tpc = 8

s = r / tpc = 4 / 8 = 1/2 serpentine

The problem with serpentine is that you can't just pick any ratio and draw the circuiting. For example - given a 4 row deep coil you can't just draw a 3/7 serpentine circuit. But you can draw a circuit for any tube per circuit count (obviously between 1 and the total number of tubes in the coil).

All of the above examples deal with a vertical coil. For horizontal coils, the same logic applies but you get a completely different circuit drawing.

Notice that the tubes are linked in such a way that liquid (water or refrigerant) can drain freely to the bottom and air bubble (in water coils) will rise freely out the top end.

There are very few technical design guides published in this field. The ASHRAE handbooks don't really cover the subject adequately and I haven't found much in the other refrigeration texts. My best reference is someone who has circuited coils for all of his working life. Peter Timm runs the coil shop at HC-HeatExchangers, South Africa and over the years has shared much of his design knowledge with me.

If you have comments, I would be very happy to hear from you.

In the development of this article, I have made an Excel spreadsheet that
calculates any valid circuit for any size and shape of coil. The spreadsheet clearly
shows the benefit of tpc over serpentine by tabulating every possible coil
circuiting. If you would like to get a free copy of my **circuits.xls** spreadsheet,
send me a picture postcard (a real one, not email) from you home town.

TechniSolve Software

Bruce Wernick

PO Box 1813

Randburg, 2125

South Africa

Ask for the free circuits spreadsheet and give me your name and email.

You will also get the 1st release of my new auto coil circuit software.

Copyright TechniSolve Software © 2005 All rights reserved

For more information and software, visit the web site