Many chemical, oil & manufacturing plants were originally constructed more than 30 years ago. These plants often have aging firewater distribution systems with capacities and capabilities that are not well understood by plant personnel. In many cases, due to personnel changes and a lack of up-to-date system drawings, important information regarding the firewater system layout may even be missing.
A number of factors may motivate a plant to update or develop accurate drawings of the firewater distribution system. These factors may range from maintenance problems, performance issues, an inability to isolate system sections, system additions, and concern over system adequacy in emergency situations. An improved knowledge of the firewater distribution system can be critical in both emergency and non-emergency situations.
Although improved knowledge of the firewater system can be useful, even more valuable information can be obtained through the hydraulic modeling and simulation of the distribution system. Information that can be obtained from such an exercise includes knowledge of where pressure and flow are greatest, understanding the impact of system additions and modifications and knowledge of the impact of disabling portions of the system.
Over its lifetime, a firewater system will undergo changes on a regular basis as a facility expands and changes and as repairs are made. Detailed CAD drawings and records have only been around for the past 15 to 20 years. Prior to this, paper records were the primary way of maintaining information on firewater systems. It is unusual for accurate, detailed records of older systems to exist. To develop an up-to-date picture of the existing system, existing records must be verified and the system must be evaluated in detail.
The first step in evaluating a firewater system is gathering enough information to develop an accurate system layout. This involves interviewing knowledgeable plant personnel and reviewing written records and drawings. Information gathered should include: piping layout; pipe diameter; pipe type; pipe age; pipe connections; valve, hydrant and monitor location and firewater sources. Invariably, some incorrect or missing information on an existing system will exist. Finding missing or incorrect information can be a difficult task and may require digging up sections of piping. Often, by isolating various valve combinations and flowing water from different locations, it is possible to deduce information without digging up pipe sections.
Depending on the age of the system, several different types of piping may be utilized. For older systems, cast iron piping is typical, often of 6-inch or 8-inch diameter. Other widely used types of piping include ductile iron and carbon steel. Cement-lined pipe is often used in environments where corrosive soil and/or water conditions exist. In the past 10-20 years, several types of plastic piping have been introduced and are regularly used in plants. Plastic piping is predominantly high-density polyethylene (HDPE) but may also include Polyvinyl Chloride (PVC) and fiberglass-reinforced plastic (FRP). The advantages of plastic piping include higher coefficients of friction and resistance to corrosion. Disadvantages of plastic piping include durability and concerns over exposure to hydrocarbons. Newer piping of all types is typically installed with diameters from 8-inches to 16-inches.
To develop a hydraulic model of the system, understanding the layout of the distribution system is only the first step. It is essential that accurate data on pumps and water supplies be obtained. Output data is only as accurate as the input data. It may be necessary to complete pump testing to generate a pump curve. The use of manufacturer's pump curves should be avoided. If there are connections to municipal water sources or adjacent facilities, testing should be completed to determine the available supply from these sources. In addition, elevated water supplies and piping will have an impact on system pressures. A topographical map of the area should be obtained if available.
The accuracy of a hydraulic model is dependent on having a detailed knowledge of the condition of the piping. To do this, it is not usually necessary to dig up a section of piping and observe the interior condition for incrustation or tuberculation. A knowledge of the available water supply can provide insight into the amount of deterioration over time. The use of salt water, water subject to the growth of biological organisms, or sediment-rich water may accelerate piping deterioration. Flushing the system may also provide indications of the quality of the water and the effect it will have on the piping over time.
However, there is a more accurate and scientific method of determining pipe condition. The Coefficient of Friction, or C-factor, is an indication of the roughness of the internal pipe surface. The higher the C-factor, the smoother the pipe interior. In design applications, book value C-factors are used for new pipe sections. Historical book values for piping of different ages are also available. See Table 1. Over time, the internal roughness of a section of pipe will change. Due to factors such as corrosiveness and sedimentation, the C-factor in existing pipe can vary considerably. As a result, actual C-factor testing will provide a more accurate determination of the coefficient of friction than historical values obtained from a textbook.
Table 1. Historical C-factor Book Value
C-factor testing is a method of qualifying the coefficient of friction of existing underground piping within a network. The results of C-factor testing can be used to assess the interior condition of existing pipe. The Hazen-Williams formula is the primary friction loss formula utilized for fire protection hydraulics. Other formulas such as Darcy-Weisbach and Chezy also exist. The Hazen-Williams formula normally appears as:
p= pressure loss per linear foot of pipe (psi/ft)
Q= flow (gpm)
C= Coefficient of friction
d= Internal pipe diameter (inches)
By measuring the flow and pressure drop along a known length of pipe, the C-factor can be calculated for the pipe section by rearranging the formula as follows:
L= Lenght of pipe section between the pressure readings (feet)
p1= Pressure at gauge 1 (psi)
p2= Pressure at gauge 2 (psi)
The actual internal pipe diameter will not be the same as the nominal pipe diameter shown on firewater system drawings. The internal pipe diameter can be obtained from the pipe manufacturer or hydraulic textbooks. Using the listed nominal pipe diameter will result in an inaccurate calculation.
To properly conduct a C-factor test, the piping network must be configured such that the water flows in one direction through the pipe section being tested. To accomplish this in looped sections, one or more valves in the piping network must be closed to ensure water flow in a single direction. If the reliability of closed valve(s) is in doubt, the feed line should be isolated prior to the test and a flow connection opened to ensure that the valve(s) is functional and no unknown lines are feeding the test section. A 'no flow' test result will serve as verification that the valve(s) is functional and the test section can be isolated.
An optimal section of piping for testing is a section of the same size and type of pipe with as long a distance between the gauge points as possible. The greater the pressure differential between the gauge points, the more accurate the test results will be. If only short pipe runs are available, flowing large quantities of water will increase friction loss and improve the accuracy of the test. For each test, there must be two gauge points and at least one flow point.
The following procedure can be used to complete C-factor testing. Figure 1 below provides a diagram of a sample test section.
C-factor Testing Procedure:
Verify that the available water supply or pump can provide adequate volume and pressure. To obtain a meaningful pressure drop between gauge points, it is often necessary to flow large quantities of water.
Confirm that any valves needed to properly isolate the section being tested are working properly. Close the valves.
Install pressure gauges on P1 and P2 (typically fire hydrants) and determine static pressures.
Open the flow point (Q) and obtain a pitot reading.
Obtain and record pressure reading P1 and P2 from pressure gauges. Multiple flow points maybe needed to flow larger quantities of water and if a reasonable pressure differential is not obtained between P1 and P2 .
Slowly close valves to flow point Q and pressure gauges P1 and P2 .
Reopen any closed valves from Step 2.
Remove gauges and any nozzles used from P1, P2 and Q.
Determine distance between gauge points P1 and P2. This should include equivalent pipelengths due to fittings and valves. It may also include equivalent pipe lengths due to a need to test pipe sections where different diameter piping is used.
Figure 1. Sample C-factor test.
As with any water flow testing to be completed, safety precautions should be taken to avoid property damage. Elbows, diffusers, stream straighteners and other tools should be used to manage the flow of water and assure that accurate water flow data is obtained.
Once the test has been completed, C-factors should be determined using equation 2 above. These C-factors can be compared to the historical bookvalues from Table 1 to make sure that wild inaccuracies do not exist.
At this stage, all necessary information to create a hydraulic model of the firewater distribution system has been obtained. An updated system layout will have been compiled including information on pipe diameters, lengths, type and coefficient of friction. The only remaining obstacle is putting the information in a framework that allows for specific questions about system performance to be answered. Due to the complexity of most firewater distribution systems, the only practical way to evaluate a system is to use a hydraulic modeling software package. Hydraulic modeling software enables the user to input all the relevant data concerning the water distributions system. Depending on the software package used, the model can be manipulated to provide valuable information on the performance of the system.
There are a number of software packages available to complete hydraulic modeling. These packages include EPANET 2, PIPE 2000 and WaterCAD. EPANET is a free hydraulic modeling software package, which was developed by the Environmental Protection Agency. Although the EPANET software was specifically developed to help water utilities maintain and improve the quality of water delivered to consumers through their distribution systems, the additional functionality offered by some of the commercial modeling packages may be worth the investment.
Once the hydraulic model has been created, a variety of 'snapshot' simulations can be run to determine the functionality of the system. Variables that can be modified to gain a better understanding of strengths and weaknesses of a distribution system include demand volumes, demand locations, pipe sizes, connection points, valve locations, pumping capacity, pump location and available water supplies. A wealth of useful and important information can be gathered including available flow and pressures, areas of high friction loss and the impact of system improvements.
An example of applying a fully developed hydraulic model is to analyze a process area with a high firewater demand. The high demand could be due to deluge systems, cooling water requirements, large hold-up volumes or a combination of these. The hydraulic model would first be used to determine if an adequate volume and pressure is available at the expected demand points. Should inadequacies exist, a cost-benefit analysis of different options could be completed. Potential options may include additional pumping capacities, connections to municipal or neighboring water supplies, additional piping, replacement piping or connections to improve the system grid. Each option that provides satisfactory results can be compared in terms of implementation costs to determine the most cost effective option.
Many facilities have learned to live with a firewater distribution system that is not fully understood. Although a potentially time-consuming task, the system can be hydraulically modeled. This will result in a thorough understanding of a systems capabilities and capacities. In addition, a hydraulic model can be used to evaluate multiple options for system improvements to determine the most cost-effective solutions.