1. Velocity used to be in NFPA 13. Originally, most of the data was imperically derived. I think the requirement was there because they really didn't have data to support high velocities, and they didn't know how it could affect the pressure loss in the pipe.
2. I have been calculating fire sprinkler systems since 1983. I rarely start with a pipe schedule system, mainly because the pressures where I work vary substantially. I can have static pressures from 50 to 120 PSI, literally where one zone ends and the other begins across the street. Experience is the best way to determine starting pipe sizes.
That being said, start with something - pipe schedules can be great, but don't vary feed main or crossmain sizes. Use pipe schedules as a starting size for branchlines.
Look at the flow from each outlet - one of them will be driving the calculation. Try to keep the other outlets from over-flowing too much. Often, the answer is to decrease the pipe size to an over-flowing area - rather than to just start increasing pipe sizes. Sometimes your changes will move the remote area - so you may have to do more calcs.
Make sure you are starting with the right orifice size. Do not use a 5.6k head for high piled storage densities. Don't always used a 5.6k head for attics either. At 7 PSI, in an area that has a demand of .1 gpm/sqft, a 5.6k head will cover an area of 148 sqft. It is required to flow that much, even if the area you need to cover is less. IE; If you are protecting an obstructed combustible attic with a maximum coverage area per head of 130 SqFt - drop down a size and use a 4.2k head.
Look at the total flow through the system - is it within an acceptable range relative to the minimum required flow? For example, if you have a remote area of 1000 SqFt, and a required density of .1 gpm / SqFt - the minimum flow would be 100 gpm. I would be very concerned about a design that requires 175 gpm. This indicates an out of balance system. It also requires a larger supply main to provide 175% of the minimum required flow.
If a system looks unbalanced, check the outflow from each orifice first, then check the velocity through your piping system. Pipes with higher velocity will have a higher presure loss. Sometimes that is ok, for instance I would expect a small pipe right off the main, feeding a single hydraulically closer outlet to have a high velocity. In that case you want to reduce the pressure available to keep the outlet from overflowing. Balance the flow of each outlet to the minimum demand required.
Increase pipe size to large orifice sprinklers - including drop/sprig sizes. Don't normally try to force 40 gpm through 1" pipe, even for a couple of feet - it will kill your main sizes. There are always exceptions, so play with it a bit.
Another way to balance your system outflows is to have the most remote heads covering a smaller area. If you have five heads on a line, space them so that the head closer to the main has the highest flow requirement. NFPA 13 allows you to calculate the area of coverage for each head.
Remember that on hydraulically designed systems designed per NFPA 13, small rooms with one head such as bathrooms and closets can be ommitted from flowing in calculations. Turn them off. If you have 8 closets in a row, cover them with a single 1" branchline. Turn off double compliment sprinklers, and sprinklers under obstructions such as ducts. Don't flow more outlets than what are required. The trick is almost always in the exceptions.
Unless your system is laid out typically you should calculate the required minimum flow from each sprinkler. Use the SxL method unless light hazard and small room rules both apply. In that case take the room divided by the number of sprinklers to figure the area of coverage for each head. This is important - ideally, an 800 sqft room, meeting small room requirements at a density of .1 gpm / sqft, should require 80 gpm. If the room lay-out requires 5 heads, flow 16 gpm per head - not 22.5 gpm per head. Wall distance and distance to next head or line do not always matter in small rooms.
Remember that outlets with the same k-factor that are hydraulically closer to the supply will flow more than more hydraulically remote outlets. Modify your system lay-out, pipe sizes, head count, or anything to balance the required demand with the pressure available at an outlet.
The one thing you can't do is mix orifice sizes on closed head systems in the same area to balance calculations. You can and should vary orifices on deluge systems, but that's a whole other "Oprah."
Good Luck!
