Archimedes screws (strictly speaking, hydrodynamic screws or reverse Archimedes screws) are not the most efficient technology, but they are simple, robust, and fish friendly. If we used a more efficient technology such as a Kaplan turbine, we would have to add fish screens, and they would need automatic cleaning equipment, and it would become too complex and expensive to justify.
The bigger they are, the more water flow they can take. But there is a limit. If the radius is bigger than the distance the water falls, either part of the top end will be always be in air, or part of the bottom end always in water. That would reduce the efficiency.
At Abingdon the biggest diameter we can use is about 3.4m. That will pass about 5 tons of water per second, which could generate about 60kW of electrical power. To get more power, add more screws. However as in all engineering, there are tradeoffs. More screws give a diminishing return - put in a dozen screws and most of them will not turn because there will not be enough water flow.
The maximum number of screws worth installing is three, generating up to 200kW. The flow statistics show that the average power over a typical year would then be about 100kW. However.... the feed in tariff changes the calculation. The feed in tariff is in effect a tax on fossil fuel power, that is used to stimulate investment in renewable power. However the amount paid depends on the power rating, and when it goes over 100kW the feed in tariff is reduced.
So we can reduce the capital cost by installing just two just screws, with a maximum output of 100kW total. The average over a typical year is then about 55kW. When the feed in tariff expires, in 20 years, the total output can be upgraded to 120kW, so we will be better able to pay our way on electricity sales alone.
We are not allowed to take all the water flow. There are other parts of the river, such as the Swift Ditch, which also need flow to stay healthy. So the Environment Agency specifies that we have to monitor the upstream water level, at the footbridge. If the level goes below the navigation limit we are not allowed to operate.
So how much power?
Sorry if you don't like graphs....
Day's weir has a similar flow to Abingdon, and their flow records go back 70 years. The graph shows a calculation of what the output would have been with that much flow. To get the income, multiply by £240 (the sale price plus the feed in tariff, times 1000). It includes a correction, to allow for the fact that we will not be able to generate if the Environment Agency allows the water level to go too low, and the 70 year data does not tell us when that would have happened. However it does not allow for generation stopping for any other reason.