The set up

The set up
5.46mm jet delivering 0.68 l/s to the pelton which is rotating at 900 rpm and generating 135 watts into the grid.

Monday, 25 April 2016

A new approach to dry season running

Changing nozzles to track diminishing flow can, at this time of year, be something which needs to be done rather frequently.  Last year between May and October I needed to change nozzles 11 times.

All 'direct-to-grid' connected hydro turbines need to operate at a fixed rpm: the frequency and voltage of their ac output demands it.  But a Powerspout is different because its output is fed to the grid via a rectifier and an inverter. These 'condition' the electricity produced by the SmartDrive to turn it into perfect mains quality electricity, and a consequence of being grid connected in this way is that operating at a fixed rpm isn't necessary.  However 'dirty' the output from the SmartDrive, meaning whatever the frequency and voltage, both of which are determined by rpm, the rectifier will 'clean' it up and the inverter will make it indistinguishable from mains ac. 

This being so, changing nozzles to keep track of changing flow becomes less of an imperative. The one purpose in doing it is to keep head at the design height so turbine rpm will be kept at design rpm. 

If slavish adherence to a fixed rpm isn't essential for a Powerspout, why not, as flow diminishes, let the header tank empty instead of changing to a smaller nozzle? - let a new head level establish itself somewhere down the penstock: - the place will be where the flow through the nozzle(s), reduced by the reduced head, equals the inflow to the tank; and let the rectifier and inverter do their job of making good the 'dirty' electrical output caused by the pelton turning at a reduced rpm.

In short, instead of aiming to generate as a fixed head, fixed speed installation*, think instead: variable head and variable speed.

I have not yet tried this way of operating.  In a test to see what would happen if the tank was allowed to empty, the observations were supportive of the theory, so I intend to give this new approach a try. (The test is reported here in a Dec 2014 blogpost.)

Instead of the 11 nozzle changes of last year, I reckon I might get away with just 2 or 3.  The big downside will be a loss of efficiency in the system because it is operating away from its design parameters, so that for a given flow, less will be generated compared with the conventional, fixed head approach;  and there will be the downside too of silt being carried into the penstock from the bottom of the empty header tank. But on the plus side, every drop of water at this time of year when water is less will be used for generating, instead of some being lost because the header tank is full and overflowing.

We'll see how it works out ! - but I suggest others don't follow in case it leads to unforeseen disaster. I'll report back.

* with a Powerspout the operational speed is actually never quite fixed, it varies with flow / power output, so referring to it here as a fixed speed installation is slightly inaccurate. In theory, if the head is constant, the optimum speed will be constant, so in that sense it is a fixed speed installation.

2 comments:

Anonymous said...

Bill, an interesting lateral approach to running the Pelton. With the possible reduction of daily maintenance a big plus.
I am puzzled by the sudden drop of power when the tank emptied at 9-00 pm.(15 watts). I have noticed on my penstock mock-up that a vortex forms as the tank level approaches the top edge of the pipe outlet. I wonder whether the formation of the vortex causes some loss of kinetic power in the flow. The vortex seems to disappear if the penstock inlet is horizontal and a bigger diameter than the penstock. A right angle funnel one option.

Bill said...

Hi Jim (at least I'm guessing it's you signing in as anonymous!) The reason for that sudden drop is that in an instant, the head drops suddenly from being the level of the last dregs of water in the tank to a new level way down the penstock. The sudden change of head accounts for the power drop. At least that's my reading of it. A vortex can be a problem (especially funnelling air into what should be a flow of only water) but the water is moving so slowly in my scheme that one doesn't form. Thanks for commenting. Bill