It's been a while since you posted, so I hope this reply is still relevant.
I have a wee bit of understanding about fluid dynamics...I hope this helps...
Fluid flow rate in a pipe (and through a valve) is proportional (but not linearly proportional) to both area of opening and pressure. (These relationships have been under a great deal of study for a long time. There are a couple of equations out there that are beyond this discussion.) To discuss flow through a valve, therefore, requires understanding the relationship between "% open" versus "% area opening" versus pressure.
1. There is a difference between "% open" versus "% area opening."
"% open" is what we see when we move the valve operator or spindle. We believe when we turn a butterfly valve 45 degrees, it is 50% open. (A butterfly valve only has 90 degrees of movement from open to closed. So 45d/90d x 100% = 50%.)
"% area opening" is the area that the water actually "sees" available to flow through, often called the throat. As seen in this graph: http://www.spiraxsarco.com/images/re.../fig6_5_02.gif
, the area of the throat does not change proportional to the amount of valve closure.
In other words, a butterfly valve that is only 30% "open" still has 60% throat area available for fluid flow. The difficulty is that the throat area changes very rapidly in the last 25% of closure.
2. A bit of mathematics.
Now, we relate the area to fluid flow using the equation:
Flow = Throat Area x Velocity of fluid.
Given the observation of valve behavior in #1, when closing a butterfly valve there will not be a corresponding change in the amount of flow until the valve is nearly closed because the area of the throat stays so high throughout most of the valve's closure, making it hard to control flow.
3. Constant pressure versus dynamic pressure.
The discussion in #1 and #2 (above) assumes the valve is under constant pressure. BUT, since the throat is changing, so is the system pressure, but not as we might expect:
Let's consider a valve being closed: as throat area decreases, fluid flow decreases through the valve, causing less velocity and less friction loss in the system, so less pressure loss, therefore the pressure the pump works against decreases.
Now, if your system is hooked to a typical beer pump, it is likely a centrifugal pump (and not positive displacement), where the lower the working pressure the more
flow the pump delivers. So even though the area of throat decreases, the flow rate does not decrease as much as the throat area decreases. In essence, the pump works against the valve, compounding the problem.
4. Friction versus flow.
As stated before, the relationship between friction and flow is not linear. Plus, the relationship between working pressure and flow output (in a pump) is not linear. So, changes in working pressure will not see a corresponding changes in velocity or flow. The problem is compounded again!
5. Valve flutter.
Butterfly valves tend to "flutter" near closure because of the fluid forces on the valve. The fluttering creates variations in the throat size, and therefore variations in quantity of flow. Now the problem is even more compounded.
As you can see, flow control is a complex problem. So, the best control valves are those designed to provide a linear relationship between "% open" and "% throat area." Generally, a butterfly valve is not consider a flow control valve, and is best used as purely open or purely closed.