Issue : October-December 2002

Investigation of Condensation Problem on Chilled Water Piping

By Richard Stone
Director BDP Advanced Technologies Asia Pacific Pte Ltd, Singapore
Atul Mathur
Content Alive, Singapore

Richard Stone is a mechanical engineer with honours degree from UMIST in England. He is employed by BDP Advanced Technologies and heads-up their Singapore office. BDP AT specialises in the design and delivery of Semiconductor, Electronics and Pharmaceutical projects comprising cleanrooms and high integrity environments. He can be contacted at rc-stone@at.bdp.com.sg.

Atul Mathur is a mechanical engineer with a master's degree in Thermal Sciences from IIT Kanpur, India. He has 12 years of experience in HVAC contracting and consultancy. His firm, Content Alive, provides HVAC consultancy and technical writing services. He can be contacted at atulm@singnet.com.sg

Chilled water pipes are insulated for energy efficiency, avoiding heat gain to the chilled water and preventing surface condensation. In most cases, the insulation is installed following standard specifications as it is not expected to be an area of concern for typical air conditioning systems. However, this may not be the case when considering humid climates. Unless selected carefully and installed with attention to good workmanship, the chilled water pipe insulation could lead to problems in humid climates. This article reviews a case where problems with the chilled water pipe insulation became the centre of investigation due to surface condensation.


The project where the condensation problem occurred is at an industrial facility in Singapore. The central chilled water plant, comprising 4 nos. 2,500 TR chillers, provides chilled water to serve process and comfort air conditioned areas. The central plant is installed in the Utility Building, which is separate from the manufacturing block requiring the chilled water. The chilled water piping passed through external areas, exposed to outside conditions and internal areas.

The pipework insulation material used for this project was a cellular type glass insulation. The specifications called for the insulation material to be applied to bare pipes with a coat of mastic, and then, strapped using stainless steel bands and finally, covered with stainless steel jacketing. The key properties of the insulation material used were as follows:

Condensation on the outer surface of insulation jacketing occurred when the plant was commissioned and the chilled water system brought into operation. The condensation, however, was not present over the whole system, occurring only on certain sections of straight pipework, at fittings and at the pipework supports.. The condensation in some cases was not consistent and appeared only at certain specific times of the day.

The surface condensation was classified as a problem and the authors of this article were invited to investigate the problem.



The investigation of the condensation problem was carried out by adopting the approach as shown in Figure 1. This involved running parallel activities in the following three key areas:

  1. Fibrous Insulation: Composed of small diameter fibers which finely divide the air space. The fibers may be perpendicular or horizontal to the surface being insulated, and they may or may not be bonded together. Silica, rock wool, slag wool and alumina silica fibers are used. The most widely used insulation of this type are glass fiber and mineral wool.
  2. Cellular Insulation: Composed of small individual cells separated from each other. The cellular material may be glass or foamed plastic such as polystyrene (closed cell), polyurethane, polyisocyanurate, plyolefin, and elastomeric.
  3. Granular Insulation: Composed of small nodules which contain voids or hollow spaces. It is not considered a true cellular material since gas can be transferred between the individual spaces. This type may be produced as a loose or pourable material, or combined with a binder and fibers to make a rigid insulation. Examples of these insulations are calcium silicate, expanded vermiculite, perlite, cellulose, diatomaceous earth and expanded polystyrene.
I. Field observations

The field observations were focused on the following three issues:

II. Review of project specifications and manufacturer's technical literature
III. Theoretical analysis

Theoretical analysis, involving calculations for pipe surface temperature based on various parameters, such as the K value of insulation material, ambient conditions, etc., was carried out to check the following:




The findings of the investigation fall into three key categories: field observations, project specifications and theoretical analysis.

Field Observations – Temperature Measurements and Areas of Condensation

Measurements of the chilled water insulation surface temperature and prevailing ambient conditions (dry, wet bulb and dew point temperatures) were carried out at six different locations. Two locations were external and four internal. The temperature measurements were carried out over a period of four days and were varied to allow measurement during different hours of the day to produce readings around the clock (including late night and early morning hours). The measurements were carried out at specific points where problems had been identified i.e. straight lengths of pipe, pipework supports, flanged joint, elbow and valves, etc.

Table 1 shows the summary of the measurements for the prevailing ambient conditions.

Table 1 Summary of prevailing ambient conditions based on site measurements
Dry Bulb
Temp. (°C)
Relative Humidity Remarks
External areas
range not relevant*
range not relevant*
range not relevant*
9 out of 25 occurrences (max. simultaneous)
12 out of total 25 occurrences
2 out of total 25 occurrences
Internal areas 28.1-29
range not relevant*
range not relevant*
range not relevant*
8 out of total 31 occurrences (max. simultaneous)
16 out of 31 occurrences
6 out of total 31 occurrences
2 out of total 31 occurrences

* It was identified that ambient dry bulb temperature was not relevant for resolving the condensation problem and the effect was driven by the relative humidity. This is discussed later.



It is important to note from Table 1 that the relative humidity was in the range of 76-80% most of the time, but it could peak at 90% on certain occasions.

The pipe surface temperature measurements revealed the following:


Having carried out the insulation surface temperature measurements, it was found that although the same insulation material and thickness was used throughout, the insulation on the pipe fittings was providing lower thermal performance as compared to the straight runs of the pipe. The reason for this was identified as a workmanship defect and is explained in the next section.

Field Observations – Workmanship

Chilled water pipe insulation was removed from select locations to review the workmanship of the installation. The inspection revealed that there were deviations from the manufacturer's installation guidelines. The key ones are mentioned below:

Most of the workmanship issues were manifested at the pipe fittings, such as valves, elbows, flanges, etc., where specifications called for an elaborate arrangement.

Review of Project Specifications

The review of project specifications and the manufacturer's technical literature revealed the following three key issues regarding design ambient conditions, insulation thickness and insulation jacketing:

Design Ambient Conditions

Insulation specifications, in particular, did not indicate the design ambient conditions. However, the design was based on the following outdoor conditions mentioned elsewhere in the general air conditioning specifications and the same were also assumed to apply to the insulation material:
Dry Bulb Temperature: 33.3°C
Wet Bulb Temperature: 27.8°C

The manufacturer's technical literature, however, qualified the design ambient conditions as follows:

"Designers are cautioned that calculating average ambient conditions for a particular location may provide quite reasonable insulation thickness values, but this is likely to result in a significant number of days during which sweating and dripping of condensate occur."



The above statement is significant as it is not uncommon to state design outdoor conditions in the general section of air-conditioning specifications and omitting the same information in the section on insulation specifications.

Table 2 : Thickness of insulation as per project specifications and manufacturer's recommendations
  Pipe size/Insulation thickness (mm)
  100 NB 150 NB 200 NB 250 NB 300 NB 400 NB
45 45 50 50 50 50
(90% relative humidity)
63.5 63.5 63.5 63.5 76.2 76.2

Insulation Thickness

Manufacturer's technical literature provided recommendations on the thickness of insulation for different ambient humidity levels. The thickness of insulation specified in the project exceeded the values recommended in manufacturer's literature at ambient relative humidity level up to 80% but fell short of recommendation for 90% humidity. Table 2 shows the difference between thicknesses specified for the project and the manufacturer's recommendations.

It is evident that the designers did not follow manufacturer's recommendations and compromised the thickness of insulation for 90% and above humidity levels.

Metal Jacketing

Project specifications required stainless steel jacketing to be applied over the insulation. The manufacturer's literature, however, cautioned against the use of highly reflective materials as follows:

"On piping and equipment with operating temperatures below ambient, highly reflective materials with low emissivity, such as unpainted metal jacketing, will decrease heat gains. As a result, the surface temperature will be reduced and the potential for condensation will increase. When designing below-ambient insulation systems for maximum condensation protection, less reflective materials with higher emissivity such as painted metal, PVC, PIB sheet, ASJ or mastic should be selected for the outer surface of the insulation system."

This is again an area often overlooked while specifying the type of jacketing over chilled water piping insulation.

Theoretical Analysis

In addition to the field observations and review of project specifications, a theoretical analysis was carried out to check the following:

The analysis involved using the following formulae for steady state heat transfer for a hollow cylindrical geometry:
(1)     q = U A ΔT
where, q: Rate of heat transfer (W)
U: Overall heat transfer coefficient W/(m2•K)
A: Area based on inner radius of pipe (m2)
ΔT: Temperature difference in K (ambient dry bulb temperature – chilled water temperature)


hi : Convection heat transfer coefficient at inner surface of chilled water pipe [W/(m2•K)]
ho : Convection heat transfer coefficient at outer surface of insulated chilled water pipe [W/(m2•K)]
r1 = Inner pipe radius (m)
r2 = Outer pipe radius (m)
r3 = Outer radius of insulated pipe (m)
Kpipe = Thermal conductivity of pipe material [W/(m2•K)]
Kinsulation = Thermal conductivity of insulation material [W/(m2•K)]

(3)     q = (T2 – T1) / R1 = (T3 – T2) / R2

T1 : Temperature at the inside surface of chilled water pipe
T2 : Temperature at the junction of outer surface of chilled water pipe and inner surface of pipe insulation
T3 : Temperature at the outer surface of chilled water pipe insulation
R1 : Thermal resistance of pipe material
R2 : Thermal resistance of insulation material


The above formulae were set-up on an Excel spreadsheet to calculate the chilled water pipe surface temperature for different conditions. All the calculations were performed for a sample case of 350 mm diameter pipe (as found on site).


Results of Theoretical Analysis

In order to carry out the theoretical analysis, the starting point was to establish the most challenging ambient conditions. Using the meteorological data recorded for Singapore for the period Jan 1981 to Dec 1999, it was concluded that the conditions with humidity levels ranging from 86% to 90% were the most challenging. Although, there were occurrences of humidity levels above 90%, such conditions were deemed to be close to precipitation, and therefore not considered. Moreover, when conditions with humidity levels above 90% occurred, there remains a very little gap (~ 1-1.5 °C) between the ambient dry bulb and dew point temperatures. Selections based on tolerances as tight as this would always be subject to condensation as small fluctuations in operating parameters and ambient conditions (wind, emissivity, etc.) could easily lead to pipe surface temperature dropping below the ambient dew point temperature. For the purpose of analysis, 90% relative humidity was considered as the limiting condition. If condensation occurred at 90% relative humidity, then one could conclude that it would also occur at humidity levels above 90%.

The key figures set-up on the Excel spreadsheet for the theoretical calculations are as shown in Table 3.

Table 3 Ambient conditions for theoretical analysis
Ambient dry bulb
temperature °C
27 26 25 24
Ambient wet bulb
temperature °C
26.6 25.7 24.7 23.7 22.8
Ambient relative
humidity %
90 90 90 90 90
Ambient dew
point °C
26.2 25.2 24.2 23.3 22.3

The calculations for chilled water pipe surface temperature were carried out for the limiting condition of 90% relative humidity at ambient dry bulb temperatures ranging from 24 °C to 28 °C.

The other input parameters are as below:

The results of the calculation are summarised in Table 4.

Table 4 : Summary of results from theoretical analysis
Ambient dry bulb
temperature °C
27 26 25 24
Ambient relative
humidity %
90 90 90 90 90
Ambient dew
point °C
26.2 25.2 24.2 23.3 22.3
Pipe surface
temperature °C
26.1 25.2 24.3 23.4 22.4
Condensing (Yes/No) Yes Yes No No No

As revealed by the results in Table 4, the existing insulation was not sufficient for preventing condensation at 90% relative humidity levels.

In order to find out the required thickness of insulation to prevent condensation, the calculations were repeated with insulation thickness increased from 50 mm to 75 mm and the results of that calculation are shown in Table 5.

Table 5 : Summary of results for increased thickness of insulation
Ambient dry bulb
temperature °C
27 26 25 24
Ambient relative
humidity %
90 90 90 90 90
Ambient dew
point °C
26.2 25.2 24.2 23.3 22.3
Pipe surface
temperature °C
26.8 25.8 24.9 23.9 23
Condensing (Yes/No) No No No No No
Required insulation
thickness (mm)
75 75 75 75 75

As reflected by the results in Table 5, condensation could be narrowly avoided if the thickness of insulation was increased to 75 mm. Since, the increased thickness of insulation would help to prevent condensation under the most challenging conditions (90% relative humidity), it would be sufficient for avoiding condensation under the less challenging conditions (with relative humidity levels below 90%).


The findings of the investigation can be summarised as on the following page:

  1. The insulation material was specified without due consideration to the most challenging ambient conditions (90% humidity in the present case) applicable to external or mechanically ventilated areas.
  2. Manufacturer's recommendations were not followed, firstly for the recommended thickness of insulation and secondly, for incorporating reflective jacketing over the insulation without compensating for the loss of radiant heat.
  3. The installation was not carried out as per the specifications and the general workmanship was poor. As a result, most of the pipe fittings were experiencing heavy condensation.
  4. Important issues for consideration.
    1. Standard insulation specifications should be reviewed to ensure that specific clauses are included for installations in tropical climates i.e. qualifying the assumptions for base specification.
    2. Designers should ensure that installation quality checks are carried out on site to avoid problems associated with poor workmanship. This is often missed as the insulation jacketing hides the evidence.
    3. Particular attention should be paid to detailing the pipework supports and insulation of awkward fittings. If this is over complicated, the chances of failure due to workmanship are high.
    4. Insulation K value is the biggest influencing factor for thermal performance. Standard specifications should be updated periodically to reflect the advances in insulation material technology.

Chilled water pipe insulation may appear to be a standard and routine part of all air conditioning projects but overlooking it in the design and installation processes could lead to substantial problems, especially in humid climates.