What equation are you using these coefficients in? Most of the equations that use this paradigm were developed for pipe that is a LOT rougher than HDPE. Many of them are having problems with modern steel pipe, but none work very well with HDPE. I've checked a couple and their ability to predict dP in modern pipes is really terrible (generally over predicting dP by a factor of 3-6). So if the program says 30 psi/mile and I go up a size and it predicts 12 psi/mile and I install that size and get 2 psi/mile I probably would have been better off with the first pipe.
This is a topic that has a lot of confusion around it, and it is important to be careful with the terminology that we use or the cofusion can grow. In your subject line you speak of "Equivalent Length" but in the body of your post you ask for "resistance coefficients (K)". It is important to distinguish between these two concepts when we are comparing different materials of construction.
The head loss through a fitting is caused mainly by changes in velocity and flow direction of the fluid. The roughness of the internal surface of the fitting has very little influence in normal commercial fittings and valves - regardless of whether they are fabricated from metal or plastic. This means that resistance coefficients (K values) for steel fittings can be safely used for plastic fittings of the same geometry. However, equivalent lengths given in terms of steel pipe will generally be on the low side when used for plastic fittings. This is because the head loss through the fitting is being expressed in terms of something that is not directly related to the fitting (ie. in terms of the attached pipe) and the equivalent length is not purely a function of the fitting.
This topic comes up so regularly that I wrote an article expanding on it (and related matters). If you would like more detail than I have given here please read the article.
bimr has linked to a useful document published by PPI, but I have 2 comments on their data in Table 2-2 on page 174. They have unfortunately used the symbol K' to represent the Equivalent Length. It would be very easy to read these values as resistance coefficients because that is what K usually represents - and this would lead to huge errors. My second comment is that their values for molded elbows seem rather high. I have converted resistance coefficients calculated using the Darby 3-K method for a variety of fittings to equivalent lengths in a range of materials in another article. My equivalent length values are quite a bit lower than the PPI values.
To clarify some of the confusion, PPI is not using the symbol K' to represent the Equivalent Length or the Crane friction factor (K). One has to assume that is the reason for the use of a different symbol K' than K. To use the PPI method to calculate the equivalent length, you multiply K’ times D (pipe bore diameter in ft) to calculate the equivalent length.
"Fluids flowing through a fitting or valve will experience a friction loss that can be directly expressed using a resistance coefficient, K’, which represents the loss in terms of an equivalent length of pipe of the same diameter. (20) As shown in the discussion that follows, this allows the loss through a fitting to be conveniently added into the system flow calculations. Table 2-2 presents K’ factors for various fittings.
Where a pipeline contains a large number of fittings in close proximity to each other, this simplified method of predicting flow loss may not be adequate due to the cumulative systems effect. Where this is a design consideration, the designer should consider an additional frictional loss allowance, or a more thorough treatment of the fluid mechanics.
The equivalent length of pipe to be used to estimate the friction loss due to fittings may be obtained by Eq. 2-9 where LEFF = Effective Pipeline length, ft; D is pipe bore diameter in ft.; and K’ is obtained from Table 2-2. (2-9) LEFF = K’D Fluids flowing through a fitting or valve will experience a friction loss that can be directly expressed using a resistance coefficient, K’, for the particular fitting.
Thanks for the clarification bimr - there I was pointing out that we need to be careful with our terminology and my sloppiness causes more confusion. What PPI have called the Equivalent Length, using the symbol LEFF, is a dimensional value and is the length of pipe that would be equivalent in pressure drop to the fitting. However, common (and lazy) practice is to call the dimensionless ratio LEFF/D the equivalent length, even though it is actually a ratio and not a length at all.
An advantage of working this way (i.e with the LEFF/D ratio) is that it allows us to work in whatever length units we like. If you express your pipe bore in ft you will get the equivalent length in ft, while if I express my bore in mm my equivalent length will be in mm and be just as valid as your value. It also allows us to use a single equivalent length (ratio) for all sizes of a particular fitting.
