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.

Sunday, 31 December 2017

System efficiency at top end flows

December has been a better month, - we've had 118 mm of rain, which is double what fell in November and three times what fell in October, - and it's now working its way through for me to see a pleasing uptick in generation.  Units of energy generated in December were 312 kWh (a figure which is poor for December) and the capacity factor for the month was 56%.  At the moment the turbine is putting out 892 W from a flow of 3.26 l/s and these are heights of power and flow which I have never before aspired to.

The explanation behind such an unprecedented level of generation is that I've been waiting for a time when there was copious flow so I could run some tests to check what the system efficiency is at the top end of the range of flows I see at my site. To do this I have had to run the installation a little beyond the limits imposed by the various authorities that licence small hydros: strictly the output should not exceed 750 W and the maximum flow 3 l/s.

Previously in this blog, I presented a plot like the one below to illustrate how operating with one jet in the bottom position was more efficient than operating with both jets:



With the data I had at that time, the plot seemed to show there was a clear advantage to using one jet on the bottom - it gave an improvement in system efficiency of about 2 %, and this advantage seemed to apply throughout the flow range, although it should be noted that data points for single jet operation are sparse for higher flows.

Having copious water available at the moment, it has been possible to get data readings for single jet operation very easily. This has meant I've been able to fill in the gaps in the earlier data series and better define what happens at the top end.  The latest plot, which is based on a completely new and more refined data set, now looks slightly different:





... it will be seen that when a very large single jet is employed, one that delivers 3.12 l/s, the efficiency starts to drop off and begins to fall below the trend line for two jet operation.

I'm not sure why this should be. There shouldn't be any question of water from this bigger nozzle missing the cups on the runner: the cups are 70 mm wide and the diameter of the jet 12.2 mm. Maybe it's just that there's a lot of splash from the force of this big jet and that impairs the efficiency.

Some of the reason for the fall-off in efficiency beyond a flow of 2.5 l/s in both single and two jet operation is that shaft speed begins to rise away from the optimum speed for the runner. I have found that with flows greater than 2.5 l/s I have needed to change from the standard Type 2 rotor to the more strongly magnetised Type 2+ rotor. This keeps the shaft speed down to nearer the optimum speed, which for my site is 1000 +/- 100 rpm.  At the moment, generating as I am 892 W from 3.26 l/s, the shaft speed is 1165 rpm.  Were I to push the experimenting and deliver even more water to the runner, the downward trajectory of the efficiency curves for both single and double jet operation would fall off steeply as shaft speed began to rise despite the 'braking effect' of the Type 2+ rotor.

As it happens, I don't want to do such an experiment: the inverter I have is a WindyBoy 1200w and for continuous operation (as is the case when the inverter is coupled to a Powerspout) the output rating of 1200 W has to be down-rated to 900W.  Not wanting to roast the inverter, I think I'll just stay at 892 W.

Saturday, 16 December 2017

Grid failure

This past week has seen snow falling where I live.  However well Western Power, the electricity distribution company for our area, has done its job of cutting back overhanging trees from the power lines, snow always brings problems.  In this past week we have had repeated power outages and I thought it worth writing a Diary entry on how a grid connected Powerspout pelton behaves in such circumstances.  

When the grid goes down, the turbine will continue to receive water and the energy of that water has to have somewhere to go,- where it goes depends on whether it is an older or newer type Powerspout and on what type of inverter is handling the interface between the turbine and the grid.

... the original "grid enabled" Powerspout, the GE 400, which is the one I have, manages the situation by diverting power which cannot pass to the grid whilst the grid is down by sending it to a small heater load which is splash cooled by the spray within the wet side of the turbine casing: 


 Diversion is controlled by an electronic control board, housed in the dry side of the casing, and this regulates the power fed to the heater element so that system voltage is kept at 380 volts DC.



The necessity of keeping voltage at this level is dictated by the inverter I have which is an early one marketed in 2011. It is unable to accept voltages higher than 400v DC.  The electronic control board locks system voltage at 380 v, just below the inverter's limit and thus keeps the inverter from being damaged by over-voltage.

... later, inverters came on the market capable of seeing an in-coming voltage considerably in excess of 400 v: - the 2 kW Enasolar inverter, for example, is quite happy with 600 v.  This development in inverter technology opened the door to managing grid outage situations in a completely different way, a way which was much simpler and did away with the need for a control board.  

Instead of diverting electrical power to a load, electrical power ceases to be created at all by keeping the system in open circuit: with no load connected what happens is that system voltage rises as the pelton runner goes to its run-away speed but, crucially, no current flows.  And so long as the system voltage rises to no more than 600 v the inverter remains safe.  What becomes important with this way of managing grid outages is that a stator must be selected which is wound in such a way that even at the highest possible run-away speed possible for the site (which is determined by the net head), the open circuit voltage will never exceed the inverter's limit.

You may well ask "but what happens to all the hydraulic energy created by the pelton if none of it now finds an outlet by being turned into electrical energy?" The answer is that a lot of it fails to be translated into shaft rotational energy because at the run-away speed of the pelton much of the water passes through the pelton runner, which is moving just about as fast as the water jet, without ever hitting the pelton cups.  The water hits the casing opposite the nozzle and its energy is dissipated as heat. Some extra energy is also lost as heat in the bearings and shaft seal - a greater amount at the higher speed at which the shaft is revolving at run-away speed than is lost at normal operating rpm.

To illustrate some of this, here is a picture of a pelton at run-away speed showing how the water from both nozzles fails to be deflected in the normal way onto the front glazing because the velocities of both runner and jets are little different:


This pelton was coupled to an Enasolar 2 kw inverter which, at the time, was not connected to the grid; the run-away speed of the turbine which resulted was measured at 1440 rpm:



and because the stator of the SmartDrive had been carefully selected to be one which delivered just 0.266v /rpm in open circuit (it was a 100-14S-1P S stator), the open circuit voltage at runaway was measured at 383 v: well below the 600 v limit of the inverter.

So there have been two ways of managing a grid outage with a Powerspout pelton and both are good.  This past week with its numerous grid outages has reminded me just how "bomb proof" my system is: when the grid goes down, the turbine continues happily feeding power to its dump load and after the grid comes back on, seamlessly the turbine re-connects itself to the grid.  It's all clever stuff and it gives me great pleasure to see it in operation.

Sunday, 1 October 2017

End of year results.

Good data about small hydros is hard to come by.  By good I mean energy data which is trustworthy, granular enough to show daily output and historical enough to allow comparison of one year with another. Measures such as 'Capacity factor' and 'Availability factor' are also nice parameters to see reported.

But such data is scarce so it's always a high point for me to come to the end of my 'accounting year' and be able to present my figures.  

The 'accounting year' I use runs from October 1st to September 30th so today is the day I have been able to wrap up the figures for the past 12 months.  Presented graphically as a cumulative plot of kWh's generated the results look like this:








...and the same data presented as a daily plot showing generation on each day of the year, the results look like this:




 Both plots show data for this year (purple) and for the previous three years. 

Though the figures for hydro generation are interesting enough in themselves, many very small hydro installations will be coupled on the same premises with a small solar installation, and it's nice to be able to see how the one compliments the other.  Here then are the corresponding graphs for my 3.3 kWp PV which has been operational for only two years:






It doesn't need a magnifying glass to spot some of the comparisons to be made between the hydro and solar yields:

  • solar is hugely variable on a day-to-day basis but the cumulative plot in one year follows almost identically that of the previous year;  hydro by contrast has almost no day-to-day variation (a step change occurs only when a nozzle change is made) but the cumulative plot lines are widely divergent from one year to the next.
  • the yield from hydro does well from November to May, whilst solar does well from March to September; late October is the time when total 'domestic generation'(i.e. the daily sum of the two) is at its lowest.
  • on the evidence so far available, the total per year for my domestic generation will lie between 5800 and 7200 kWh.
This last bit of information, i.e. generation being between 5800 and 7200 kWh/year, has been a key driver in my exploring 'battery-on-the-wall-storage' at home.  Our domestic consumption of electricity is about 6000 kWh/year but because cooking (our heftiest use) always takes more power than is being generated, we end up having an annual take from the grid of about 1700 kWh.  I have been exploring whether it might be feasible to reduce this by storing domestic generation into a battery when there is a surplus, - a few hundred watts at any given moment but over several hours, - so that it can be used for cooking when the need is for 3-4 kW but for a short time only

And a 'battery-on-the-wall' offers two other possibilities: the possibility of purchasing and storing grid energy when the tariff is cheap then using it at a time when the grid tariff is high, and the possibility of having an 'uninterruptible power supply' for the whole property for times when grid outages occur.

So data collection has its uses. It can inform how best to move with the times as new developments like 'home battery storage' come on the market. The man is coming to see me about battery storage next month.  I'll report back.