[Please click through the source link above to view the entire discussion of dead-end water mains. Below is only a highlighted version.]
Does anyone know of an industry/federal-level reference for the maximum length of dead-end water mains - mains on the order of 8" to 12" in diameter?
I’m not sure I’ve ever seen any sort of specific guidance from authorities on maximum length of specific sizes “dead end” mains; however, I believe I have seen statements from ASCE/AWWA/Ten States Standards etc. indicating dead ends should in general be avoided or minimized where possible e.g. “by making appropriate tie-ins whenever practical”, and of course that appropriately sized flushing structures etc. must be provided.
While perhaps the preeminent reason for this sort of precautionary guidance is concern for water quality, I believe ASCE additionally explained in the publication, “Pressure Pipeline Design for Water and Wastewater” that dead end conditions can also result in “unexpected high pressures” (I guess as a result of a sort of superposition of pressure waves in the area of the dead-end). Perhaps these concerns would be heightened in some situations with increasing length of the dead-end, and the answer to any specific design situation might lie in what acceptable level of water quality and provisions for surge can be expected/assured for the situation.
In addition to all of the things rconner has said, there are two other reasons water mains are “looped” rather than dead ended.
The first is reliability. It is desirable to feed water to any location in the system from at least two directions. That way, during an emergency or routine shut down of one line, service can be provided from the other direction.
Second when water is fed from two dircetions to something like a fire hydrant velocities and friction losses will be lower. This saves energy and cost.
The real answer to your original question may be found by analyzing the system for the maximum required flow and residual pressure at the end of the dead end line. Usually this is the required fire flow which on most public systems will be on the order of 1000 to 500 gpm and the minimum pressure is usually required, by health standards, to be 20 psig.
Thank you EVERYONE for the responses to the question. The case I have in mind concerns a campus with difficult terrain crowded with structures.
I understand the water-supply principles behind the avoiding of dead-end mains. What puzzles is the lack of a scientific and quantifiable basis for just how long a dead-end main can be, according to the various contingencies.
We all know that dead-ends are not desirable, and should be avoided, but at what point are they unacceptable to the civil engineer? And why?
If the client asks, “Can we make that dead-end main 50 feet longer?”, and I say “No”, he asks “Why?”.
Apparently the only answer to the question is in terms of policy and rules-of-thumb, not a reasoned industry standard supported by calculations.
Thank you again for the responses. If further information presents itself, it would be much appreciated.
As I tried to say in my previous post; you can calculate the maximum length of a dead end main if you know the size of the main, the flow desired or required at the end of the main and the lowest residual pressure allowed at the beginning of the main where it tees into another main.
For example if you have a 200’ long dead end main, a normal static pressure of 65 psi where the main tees into another main and you know the flow required (usually the fire flow) is 500 gpm and the lowest permitted residual pressure is 20 psi you can calculate the head loss in that main for 500 gpm. If that head loss is less than 45 psi (65-20) then you know the main can deliver at least 500 gpm. If the head loss is more then the main can only deliver something less than 500 gpm and probably needs to be up-sized. Do the math and you’ll get a “feel” for how it all works.
I think that a single 8"-diameter, 2000-foot-long, water pipe serving only a small building located at the end of the pipe would qualify as a “dead-end main”.
Though water drawn from the end of the main is constant, the demand is small in proportion to the volume of water in the main. This raises water quality issues, even if there is sufficient fire-fighting pressure.
Does anyone have a definition for “dead-end main”? There doesn’t seem to be agreement on the exact meaning of the term.
I along with our regulating authorities consider a dead end main to be one that is not looped. If there isn’t a user at the end of the line, why is it being installed “and” placed on the system in the first place. I don’t know many clients that just spend money to install waterlines to nowhere.
The main issue is water quality. You will have to perform the hydraulics to size the line for the existing or future flows. The latter is what can cause quality problems. If the line is sized for a large future flow; but, a low flow will exist for a time, the water can become stagnant.
As rconner stated in the last response, the water will become stagnant in an unlooped line if it is not flushed from the line (by users or a flushing hydrant) in a timely fashion. Stagnation can cause low chlorine residual, bacteria, odors, colors, etc. depending on the type of water being placed into system.
Looping the line will allow it to transfer water from one part of the system to another and potentially keep it fresher. Running two lines side by side from adjacent locations and tying them together at the end will not improve the quality as the water is from the same location and will not move through the pipe except when the users on that line draw water.
There is no scientific calculations to determine the length that I know of; but, often I will determine how much water the pipe will hold and based upon the existing flow see how many days storage are in the line. Then depending on the estimated time that the line will be in low flow condition, I will advise the client that they will have quality problems. They may decide to institute a flushing program for that line (even if its based upon complaints) or they can decide to install a smaller line now and upsize it later when the high flow conditions arise
Flushing is appropriate and absolutely necessary in these cases. However, most cities hesitate to accept these partially constructed loops into a city system because the city is then required to expend the time and money to flush the mains. The developer generally can’t be relied upon to do the flushing. And any customers connected to the temporary dead end main are at risk of health problems from the stagnant water. When those customers get sick, they call and complain to the City, not to the developer who constructed the line. A better solution is to install the loop and to reimburse the developer later when the next housing phase is constructed.
Returning to the original question, I stumbled across this in the 2003 edition of NFPA 1141:
“9.2.7 Dead-end mains shall not exceed 183 m (600 ft) in length for main sizes less than 250 mm (10 in.) in diameter.”
This is an industry standard as regards fighting fires. Any other industry/federal-level standards would be appreciated.
It appears this thread has found perhaps the most notable or detailed guidance on this basic subject to focus on fire protection (that was of course not specifically mentioned in the intitial inquiry). In this specific regard, I believe that there is now also a relatively new (first edition 1989, but now in its third printing) AWWA manual “M31”, “Distribution System Requirements for Fire Protection”. It appears this roughly 60 page manual highly stresses the importance of “dual” (providing both water supply and fire protection) system “reliability”, and I noticed the first thing it mentioned when it talked about “out of service conditions” was “loops”. While not an issue in your case but perhaps of general interest of those reading this thread, I noticed it also states (on the very first page of Chapter 1 “Fire Flow Requirements”) , “The most significant aspects are: installing and maintaining fire hydrants, providing adequate storage capacity, and meeting requirements for minimum pipe sizes (e.g. 6-in. [150-mm] pipes in loops and 8-in. [200-mm] dead ends) in neighborhood distribution mains when much smaller pipes would suffice for delivery of potable water only.”