bobc
10 kW
Thought this might be of general interest; I've been looking at a potential micro hydro project where an archimedes screw would directly drive the generator (e.g. one from a broken petrol generator), and how it could be controlled. The petrol engine speed was governed by a vane in its cooling fan airflow directly actuating the throttle. So we need a similar control for the screw turbine.
The screw is 1m diameter and 4m long - with 30degrees incline this can output about 5kW at 1 rev/s.
The main control problem is the delay from the control input (water flow control into the screw) to the output torque changing; it takes 4 seconds for water to travel through the screw.
A safety issue is that if control is lost the screw could spin faster and faster until it breaks - this sort of rules out electrical controls. So I'm looking at a centrifugal governor. For simplicity's sake I can do proportional only.
The key to making it work is to have a disturbingly large flywheel at the generator. I can't see another way to supply a sudden demand for 5kW for 4s while the screw fills up along its length....
I simulated the system in XL - looks as though it works OK. I get decent performance with a 1kgm^2 flywheel on the genny
The sim starts with 4kW output, dropping to 200W halfway through
the first graph is speed (supposed to be 1 rev/s) - so we're seeing +/- 3% or so speed. Increasing the controller gain reduces this but it soon becomes unstable. Gain can only be increased if the flywheel is made bigger.
The graph on the right is "how full are the segments" within the screw - max 69litres (0.069 m^3), min is zero
the graph below is screw output torque in Nm
I'm itching to get on with this - should be a proper giggle!
The screw is 1m diameter and 4m long - with 30degrees incline this can output about 5kW at 1 rev/s.
The main control problem is the delay from the control input (water flow control into the screw) to the output torque changing; it takes 4 seconds for water to travel through the screw.
A safety issue is that if control is lost the screw could spin faster and faster until it breaks - this sort of rules out electrical controls. So I'm looking at a centrifugal governor. For simplicity's sake I can do proportional only.
The key to making it work is to have a disturbingly large flywheel at the generator. I can't see another way to supply a sudden demand for 5kW for 4s while the screw fills up along its length....
I simulated the system in XL - looks as though it works OK. I get decent performance with a 1kgm^2 flywheel on the genny
The sim starts with 4kW output, dropping to 200W halfway through
the first graph is speed (supposed to be 1 rev/s) - so we're seeing +/- 3% or so speed. Increasing the controller gain reduces this but it soon becomes unstable. Gain can only be increased if the flywheel is made bigger.
The graph on the right is "how full are the segments" within the screw - max 69litres (0.069 m^3), min is zero
the graph below is screw output torque in Nm
I'm itching to get on with this - should be a proper giggle!