Watertransportation and distribution networks require hydraulic calculations to determination the flowrate and pressure characteristics at one or several consumption points and the water supply flowrate and pressures needed to meet the design requirements.[1]
In the context of fire safety, hydraulic calculations are used to determine the flow of an extinguishing medium through a piping network and through discharge devices (e.g., nozzles, sprinklers) to control, suppress, or extinguish fires.
The probable intensity and extent of a fire inside the building are indicated by factors including the building use, the building height, the items contained inside the building and their arrangement. These variables are compared to tables and values expressed in the model codes. The values in these tables are based on fire tests and loss history.
The water available is often determined by means of a water flow test, in which one or more fire hydrants are opened and the water pressures and flowrate are measured. Some municipal water jurisdictions may provide an estimate of available water supplies based on hydraulic models.
Suppression system piping networks are usually arranged in one of 3 configurations: Tree, Loop, or Grid. All of these types of systems utilize large horizontal pipes - "mains" - which deliver large flowrates to smaller pipes - "branch lines" - which are connected to the mains. Sprinklers are installed only on the branch pipes. The mains are supplied with water by connection to a single vertical pipe - "riser" - which is in turn provided with water by connection to water supply piping.
Looped systems utilize a main that runs a significant distance into a building and is routed back to connect to itself near the riser. Branch lines are connected to this 'loop'. Less water supply pressure is required with this looped main configuration as the hydraulic pressure drop is lower through the main as water can flow in two directions to any sprinkler. The branch lines may terminate in a dead end or may connect at each end to different (usually opposite) points on the looped main. In the latter case, less water supply pressure is required as the hydraulic pressure drop is lower in the branch pipe as water flows from both ends of the branch line to any sprinkler.
Grid systems utilizes two large mains at opposite ends of several branch lines which are connected to the mains at each end. Gridded systems provide multiple paths for the water to travel to any point in the system, reducing pressure losses in the system.
Most design standards require application of the Hazen-Williams method for determining frictional pressure losses through the piping network as water passes through it. Tree and Loop systems are simple enough that the hydraulic calculations could be performed by hand. Because hydraulic calculations for gridded systems require an iterative process to balance the water flow through all possible water paths, these calculations are most often performed by computer software. In practice, most calculations on all types of piping networks are performed by computer software. The sizes of network components can be more readily modified and recalculated on a computer than through a manual process.
The total pump efficiency, ηtotal, must be included when calculating the power input to the pump. This efficiency is the product of volumetric efficiency, ηvol and the hydromechanical efficiency, ηhm. Power input = Power output ηtotal. The average for axial piston pumps, ηtotal = 0.87.
The hydraulic motors and cylinders that the pump supplies with hydraulic power also have efficiencies. And the total system efficiency (without including the pressure drop in the hydraulic pipes and valves) will end up at approximately 0.75.
Cylinders normally have a total efficiency of around 0.95. And hydraulic axial piston motors and pumps have 0.87. Moreover, the general power loss in a hydraulic energy transmission is around 25% or more at ideal viscosity range 25-35 [cSt].
Calculate the required pump displacement from the required maximum sum of flow for the consumers in the worst case scenario and the diesel engine rpm at this point. The maximum flow can differ from the flow used for calculation of the diesel engine power.
Calculation of preliminary cooler capacity: Heat dissipation from hydraulic oil tanks, valves, pipes and hydraulic components is less than a few percent in standard mobile equipment and the cooler capacity must include some margins. Minimum cooler capacity, Ecooler = 0.25Ediesel
At least 25% of the input power must be dissipated by the cooler when peak power is utilized for long periods. In normal case however, the peak power is used for only short periods, thus the actual cooler capacity required might be considerably less. The oil volume in the hydraulic tank is also acting as a heat accumulator when peak power is used.
The system efficiency is very much dependent on the type of hydraulic work tool equipment, the hydraulic pumps and motors used and power input to the hydraulics may vary considerably. Each circuit must be evaluated and the load cycle estimated. New or modified systems must always be tested in practical work, covering all possible load cycles.
An easy way of measuring the actual average power loss in the system is to equip the machine with a test cooler and measure the oil temperature at the cooler inlet, the oil temperature at the cooler outlet and the oil flow through the cooler, when the machine is in normal operating mode. From these figures the test cooler power dissipation can be calculated and this is equal to the power loss when temperatures are stabilized. From this test the actual required cooler can be calculated to reach specified oil temperature in the oil tank. One problem can be to assemble the measuring equipment in-line, especially the oil flow meter.
This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
Are you psychic or something? I literally had to review a calc from a contractor that worked, but not very well. And I looked at the incoming service; 4" for a Extra Hazard Group 2 scenario. I looked at the pressure loss in the 4" pipe, and the pressure loss in 6" pipe. With the flow from the hydrant flow test, I put back 25 psi into the system. My other favorite is in Light Hazard, if all the sprinklers are in 2 x 2 center of tile, then the area is 196 instead of 225. Did I miss the reduction in area of the calc if the ceiling height is less than 10 feet? That's been a savior for me sometimes. Thanks for all the great input you provide! Love this site and this forum!
One big tool that was taught to me early in my career was to take the computer calculation and chart the water flow from the sprinkler to the source and account for all the pressure losses. Start pressure, elevation, backflow, and each pipe type as a separate line item. It helped with finding errors, but helped with efficiency.
When changes needed to be made, you could target the right piece. You could see if the K factor was wrong based on the start pressure. You could see if your lines were too small, or if your riser could be reduced in size without blowing up the calc. Right sizing reduces wasted water and lowers friction loss, too. Not just upsizing.
This would all be tabulated, by hand, on the computer print out and given to the design manager to check.
In addition to the bullet points above,
In warehouses I often see designers try to run grids at lengths between 150' to excess of 200' between the feed main and the float main for the overhead system. Not realizing the addition of a third leg in the middle can cut down the pressure losses through the branchline pipe over such a long runs. The thought is the savings of not having to run the additional third leg is saving money in labor and material. However, if the system is designed to where there is not a large amount of transfer lines from main to main then a third leg would be a viable option. With a third leg you could possibly reduce the feed main or riser stacks, or just all of the branchlines.
General thoughts:
Try and get water supply information any way possible during prebid. DO NOT ASSUME water supply on bid drawings is what will be provided. If its not available or unreliable take caution in proceeding with any preliminary bid design! I have seen companies try to force an owner to pay for a booster pump after they design the job and it doesn't work because they decided to get an accurate flow test after the fact.
Invest in educating designers or yourself in being able to do the hand math of basic hydraulic calculations. Do not just rely on software hoping that it will help the calcs work. Remember garbage in garbage out rule.
If possible invest in a quality hydraulic software program that really allows you to massage a design effectively, efficiently, and relatively quickly. High quality design programs do an excellent job at making this easier for those that know what they are doing when they use it.
There are other items I know that are constant issues, that I will have to add at a later time :)... but for now back to work!
Great topic, and great checklist - I'd be hard-pressed to come up with any more ideas... but I'm going to give I some thought. I was once taught that if you are scraping for every crumb, since you are accounting for the equivalent feet loss in the fittings, don't add the length of the fitting again. Avoid reducing Vics if you have to (IF you even remembered to add the RV loss). I'm finding too many designers take for granted what the computer-generated hydraulic calc or seismic brace calc tells them. Like Brian said above, examine the output data, look for potential improvements, where you are burning-up pressure; where the velocity is hurting you (e.g., second-to-the-last sprinkler on the branch line). Consider looping or bird-caging. Look where you can afford the pressure loss and massage the pipe diameters. Nothing beats the experience of doing hand-calcs, and especially with some design programs doing the calcs, I feel it is becoming a lost art. If your calcs do work, you shouldn't stop and call it good there. You need keep watching for your company's bottom line by considering most efficient in terms of material, fabrication, joining methods, and installation. What makes your company competitive. That isn't always what calcs best. That computer program doesn't know fitter rates, what length you can stuff through the trusses, and it tends to like grooved tees.
3a8082e126