[Home]	Search:
Theme: Select one Steely Old Modern Midnight Ultramarine Swiss Chocolate Traditional

Issue : October-December 2002

Selecting Cooling Coils without Proprietary Software

By Amrit B. Marathe
Partner
Opal HVAC Engineers, Mumbai

Amrit Bhalchandra Marathe is a mechanical engineer from IIT, Bombay. He has been involved in HVAC projects for more than 18 years and is a member of ISHRAE. He can be contacted at opalhvac@hotmail.com

When I started work in the HVAC field fresh out of engineering college, PCs were not commonly used in India and software programs for selecting cooling coils were rare. That was 18 years ago. Today, most large manufacturers of coils have sourced their selection software from foreign principals and this is used to select and quote for cooling coils to suit their or their customer's requirements.

There are, however, several consultants and contractors specialising in HVAC work that do not have ready and immediate access to such proprietary software. This article is meant to help them make a selection using engineering tables (based on cumbersome heat transfer equations). I have used them when I started working in this field and I have continued to use them till today. The actual performance of cooling coils in the field after installation has been satisfactory, and as desired.

Types of Coils

There are direct expansion (DX) coils and chilled water coils. Some coil manufacturers fabricate coils from 5/8 inch OD copper tubes, others from ½ inch copper tube and still others use 3/8 inch tubes. Selection of the tube size is a matter of manufacturer's choice and market demand. Price, as always, plays a major part in the tube size selection.

The selection method that I follow is for5/8 inch tubes.

Coil Construction and Geometry

In a coil, copper tubes are arranged parallel to one another, either in staggered pattern or non-staggered pattern, along the length ‘L’ of the coil. A staggered pattern is more commonly used. For 5/8 inch tubes, the triangular pitch is 1.75 inch or 1.5 inch. For ½ inch tubes it is 1.25 inch and for inch tubes it is 1 inch.

Plate or ripple fins are used to enhance the heat transfer area. Thus the primary surface area (outside area of bare copper tubes) is enhanced greatly by adding a secondary area of fins. The total area including fins is called “outside surface area,” for use in the calculations, in this article.

The cross-section (L x H) across which air flows is called the face area or the finned area. Thus L is finned length and H is fin height.

Fins are arranged perpendicular to the tubes. Fin spacing varies between 8 and 14 fins per inch of tube.

[top]

Average air velocity across the face area is called coil face velocity or simply face velocity. Thus

The number of rows of copper tubes in the direction of air flow is termed as depth of coil (rows deep). Coils with 3,4,6 or 8 rows are commonly used.

Refrigerant or chilled water enters the first row and leaves the coil from the last row. A coil in which chilled water or refrigerant is supplied to all the tubes in the first row (also referred to as tubes high or tubes in face) is called a maximum or full circuit coil. Thus a typical coil of 17.5 inch height which has 10 tubes in face (based on 1.75 inch pitch) will have a maximum of 10 circuits.

If the supply is given to alternate tubes in face, we get a half-circuit coil with 5 circuits as against 10 circuits. The U-bends at the end of the tubes can be arranged, at the time of manufacturing, to obtain the number of circuits desired. See Figure 3 for full and half circuit coils with 4 tube face.

Initial Estimate

From air conditioning load calculations, one can pick up some key values such as the dehumidified air flow (cfm), the total load (ton) and calculate dehumidified cfm/ton. This value gives a very useful clue to quickly estimate the coil cost when one is making a sales proposal. Having understood the terms such as face area, face velocity and depth of coil, Table 1 can be used to estimate coil size.

Face velocity is restricted to 500 fpm to avoid carryover of condensate from the coil. The value of 500 fpm is very commonly used for coil sizing and it works very well for cfm/ton in the range of 500 to 600. If cfm/ton ratio falls below 500, (this generally happens when room sensible heat factor goes below 0.8 due to high room latent load) a 4-row coil at 500 fpm becomes inadequate. A 5-row coil is not very common. Hence by lowering face velocity, a 4-row deep coil can be selected at 400 fpm, when cfm/ton is about 400. As cfm/ton ratio reduces further, 6-row or 8-row coils have to be selected. This situation is encountered when the occupancy and/or fresh air components are high.

Thus based on the face area and number of rows, a quick coil estimate can be done.

While actually sizing the coil, the initial assumption of face velocity may have to be changed to arrive at an acceptable selection. Thus the procedure is required to be repeated with a new value of face velocity. This is known as iteration.

Final Selection Procedure

A. Data required

For sizing the coil, the following data is required from the heat load calculations.

• Room DB temperature/Return air DB temperature (F)
• Fresh air DB temperature (F)
• Dehumidified air quantity (cfm)
• Fresh air quantity (cfm)
• Grand sensible heat factor (GSHF)
• AC load (ton)
• Apparatus dew point ADP (F) (This denotes the average outside surface temperature of the coil.)

B. Assumptions

To start with, the following assumptions will have to be made • Face velocity (fpm), based on Table 1

• Chilled water entering and leaving temperature (EWT and LWT) for chilled water coil or
• Temperature of refrigerant in the evaporator coil. (Tref)
• Number of circuits (1 ton/ circuit for " coil) • Bypass factor

In an ideal coil, we can expect 100 % heat transfer efficiency. Air passing over such a coil will leave the coil at a temperature equal to the ADP. But due to limitations in the coil construction and geometry, some air leaves the coil without getting fully cooled and dehumidified, as if a certain portion of air is bypassing the coil. The coil inefficiency is expressed as a fraction of air bypassing the coil, or the “bypass factor”. Face velocity and depth of coil have a large influence on “bypass factor”. The condition of air leaving the coil is the mixture condition of fully treated air at ADP and bypass air at the inlet condition.

We generally assume bypass factor (BF) in air conditioning load calculations. The same value can be assumed for the initial selection. The assumption can be reviewed based on the result of the initial selection (remember iteration?)

[top]

Coil geometry

1. Copper tube 5/8" OD, 22G thick, 1.5" triangular pitch,staggered from row to row
2. Aluminium fins - 8 fpi, 1.3" wide per row
3. Outside surface 22 sqft./row/face area
4. Inside surface 1.2 sqft /row/face area
5. Outside to inside surface area ratio R = 18.3
6. Metal resistance rm=0.025 (hr.sqft.F) per Btu

The metal resistance of bare pipe surface (no fins) is so small as to be of any significance compared to the normal values of other resistances to heat transfer in a coil. With fins, most of the heat must pass through the small fin section. Actual value of metal resistance for common air conditioning coils in the normal range of use is from about 0.005 to 0.03 (hr.sqft.F) per Btu, depending on the thickness and material of fins. It is not constant for different fluid conditions. But for making the selection procedure simpler, it is considered to be near-enough constant.

C. The basic equation

The equation for determining the number of rows required is as under :

where
U = overall heat transfer coefficient [Btu/(hr.sqft.F.row)]
LMTD = Log (to the base e) mean temperature difference.

D. To calculate LMTD

1. Calculate DB temperature of mixture of return air and fresh air entering the coil.
   
   (Dehumidified air = Return air + Fresh air) See Figure 4.
2. Calculate DB temperature of air leaving the coil :
   DB(lvg air) = DB(mix) – (1–BF) (DB(mix) – ADP)
   BF = Bypass factor
3. Calculate LMTD
   for DX coil :
   GTD = DB(mix) – Tref
   LTD = DB(lvg air) – Tref
   For chilled water coil :
   GTD = DB(mix) – LWT
   LTD = DB(lvg air) – EWT
   

[top]

E. To calculate U - value

The overall U-value is calculated using the following equation.

The overall heat transfer coefficient 'U' depends on the following factors.

1. Inside film coefficient ki
   Experimental data is available for chilled water and refrigerant. Please refer to Table 2 and Table 3
   ki for chilled water coil.
   1.1 Calculate water flow :
   
   1.2 Calculate water velocity through tube.
   
   1.3 Calculate k using Table 3
   For DX assume 1 ton/circuit to start with and read ki from Table 2
2. Metal resistance :
   For the selected coil geometry the metal resistance is 0.025 hr.sqft.F/Btu
3. Outside film coefficient – ko.
   Outside film coefficient ko depends on face velocity and grand sensible heat factor. Refer to Table 4

   Table 41 : Outside film coefficient (for dry coil)
   Coil face velocity fpm	ko Btu/hr.sqft.F
   100 200 300 400 500 600	4.1 6.3 8.0 9.6 11.0 12.3

   
   For wet coil (GSHF < 1) the value of ko is calculated as under :
   
   The overall U-value can now be calculated using Equation (2)

F. Calculating the coil face area :

G. Calculating the number of rows of coil :

Calculate the number of rows required using Equation (1)

The assumption of bypass factor can be verified using Table 5 based on the number of rows required as calculated above.

A safety margin of 5 - 10% is advisable.

The procedure can be repeated for:

• different value of face velocity
• different Tref or EWT.
• different ΔT for chilled water
  Please refer to the example on the next page.

[top]

What if the coil geometry is different2

• More number of fins per inch. 0
  
  
  
• ½ " copper tube coil with 1.25" pitch
  The face area and number of rows arrived at by the above procedure, gives 5 - 10% extra safety. Water side pressure drop calculations will change.
• Thinner copper tube
  The data available for 22G thick coil can be used since the effect on U-value due to change in thickness of copper tube will be insignificant. Waterside pressure drop calculations will change very marginally.

References :

1. Tested Solutions to Design Problems in Air Conditioning and Refrigeration by M. A. Ramsay. The Industrial Press 1963 (Table 2 to 4)
2. Blue Star Engineers' Handbook (Different coil geometry)
3. Voltas Engineers' Handbook (Table 5)

[top]