Mr. Fonseka's experiences

Mr. Fonseka first contacted me in November '99. Back then, he had a workshop in Sri Lanka and 10 years experience in power electronics. By March 2000, he got his first 22 kW IGC installed and working. To reach such a high capacity, he used pairs of anti-parallel thyristors (instead of triacs) and got these running without major problems. His questions to me were mainly about finding cheap generators, capacitance needed for an induction generator and the pulse transformer circuit to drive a pair of thyristors.

February 2001, he sent me photo's and a report on his 6th project. From a technical point of view, it all looked very robust and well-built. But also the number of consumers was quite impressive: 140, that should be enough for a sound economic basis. And if mr. Fonseka got so many consumers right from the start, I guess that this project must have been planned and organized very well. As far as I know, the first 5 projects were similar to this one: Single phase IGC's, up and running in a Micro Hydro project.

End of 2001, mr Fonseka moved to Canada and got a job there. As a sideline, he starts building them bigger: 3-phase IGC's and 3-phase ELC's, for customers in Sri Lanka but also in Bolivia, Spain and the Philippines.


On this page:


6-th project

On 6 February 2001, mr. Fonseka wrote:

Dear Mr. Porteqijs,

Sorry for the delay in writing to you. Though I have completed 6 projects, I could not get photograph of them. But this time I managed to get the required photographs. This is my 6th project. The site is 250 km away from Sri Lanka's capital, Colombo. This location called as "Kollonne", Managoda in Hilly area and the National Grid Point is 6 Km away from the village. The civil construction work has been done by the villages and 140 houses have been supplied with electricity by this project.

I would like to mention that this is my first 3 channel IGC and it is a success. I have done some adjustments of P and I values, other than that every thing worked perfectly. As the motor capacity is 22 kW, I used 3 of BTA40 Triacs, individually mounted on 6" x 5" Heat Sink. In addition, I used 12v DC relay mounted on the PCB to drive the consumer load contractor. All the project details are attached herewith.


Kumar Fonseka


Mr. Fonseka attached a commissioning report and a number of pictures, some of which are printed below:


The IGC housing open:

  • Top left corner: The Printed Circuit Board.
  • Top right: Three heat sinks with a BTA41B triac in the middle and a varistor over its terminals. Current rating of these triacs is 40 A, so theoretically, capacity could be as high as 230 V * 40 A * 3 triacs = 27600 W, or 27.6 kW. With 24.3 kW of dump loads connected to them (see commissioning report), they are loaded up to 88 % of this theoretical capacity.
  • Middle, from left to right, there are 2 large current transformers, "130 A Semiconductor" fuses to protect the triacs, MCB's to switch off the capacitors in case of a runaway situation, and the 65 A, 3-phase main relay.
  • Bottom: The connection terminals and capacitors to provide magnetizing current to the induction generator.

The front panel of the housing.

The tank with dump loads.

The turbine: A 4-jet pelton turbine.

The generator and transmission, with proper protection
against getting your fingers caught by a V-belt
Commissioning report, summary
Provider Renewable Energy Development Pvt Ltd
Location Kolonne - Magoda
No of Consumers 140.
Date 28/1/2001.
Generator 3-phase induction motor, 1472 RPM, PF = 0.88
IGC capacity 22 kW.
Dump loads His IGC has 3 dump load connections (or "channels" as mr. Fonseka called them). Each dump load consists of three 2.7 kW heating elements connected in parallel. So total dump load capacity is 24.3 kW.
Capacitors 1200 µF in total, in "C - 2C" arrangement.
Generator A 3-phase, 22 kW, 1472 RPM induction motor. Designed for 50 Hz grid frequency, power factor is 0.88 .
Measured data:
Gross head 43 m.
Net head 34 m.
Generator currents For the 3 phases: 53 A, 65 A and 60 A respectively
(With the turbine running on 3 jets)
Generator voltages For the 3 phases: 233 V, 230 V and 219 V respectively
Frequency without consumer load 54 Hz.
DC voltages 0.5 - 1 V DC at generated voltage, 0.5 - 1 V DC at dump load voltages


Email of 30 August 2002

Dear Portegijs,

I could not contact you for a long time. I hope you are keeping well. Recently I moved to a new house and my email address got changed since the service provider is unable to provide the service.

I have completed a project for a Sri Lankan Company. It is a 50 Kw 3 Phase IGC. The photos are attached herewith. In addition to that I have done one ELC(40KW 3 PHASE) for Philippine company. But haven’t heard from them yet. I tried to contact them many times too.




The 3-phase IGC electronics: Just 3 single phase PCB's. Transformers, relay's, fuses or circuit breakers are not visible so I guess they will be placed somewhere else in the housing.
Below the PCB's, there is one large heat sink with, for each phase, 3 triacs. In an earlier email, mr Fonseka wrote that he intended to use BTA40-600B (Package RD91) triacs. Under optimal cooling conditions, this type can conduct 40 A and then capacity of this IGC would be:
3 (phases) * 3 (dump loads / phase) * 230 V * 40 A * 0.001 (to convert W to kW) = 82.8 kW
So using it at 50 kW leaves a healthy safety margin.


Email of 2 October 2002

Dear Portegijs,

Couple of months ago I fabricated a 3 phase ELC for a Bolivian person. He installed the ELC and came across a strange situation This is what he mentioned in his letter:

At first, we calibrated the generator (AVR) for it, we have run the turbine with out the ELC and we have got 220 V. in each phase and 50 Hz. When we have connected the ELC suddenly the voltage and frequency have changed. The voltage got 320V (three hundred twenty V). and the frequency 55 Hz, immediately again we have calibrated the AVR of the generator.

I don't know why the voltage and the frequency have been changed if at first its were calibrated. Every time when we turn on the turbine we need to do this practice. Which is a little uncomfortable for the operators of turbine, because if the voltage is not regulated immediately the ELC can burn.

I can't find any relationship between the generator voltage and the ELC since the AVR controls the Voltage.
What could be the reason?




I didn't understand what caused this "strange situation" either. So I could only write back something about possible disturbance of the generator AVR by the ELC. This didn't explain all of it, as also frequency was above set point so this ELC wasn't doing its job either. Mr. Fonseka didn't mention it in subsequent emails so I guess the problem was solved.


Email of 4 March 2003


Dear Mr Portegijs,

I couldn’t contact you for a long time since I was busy with my new office work. I have stared my own company in here. Electrical panel design and building is my main activity and hydropower related activities plays supporting role. I have been getting little orders and one of my friends is supporting me to carry out the existing work. The company name is Denon Technologies Ltd. Anyway I have some thing for you. Can you recall couple of weeks ago you sent me a email regarding a ELC inquiry. I made a 40Kw elc for them and the photos are attached herewith.

Drop me a couple of words when you get a chance.





40 kW ELC, heat sink is the biggest part. Probably it still needs a fan to keep it cool...


Note: To avoid spam, email addresses are hidden by a script.
If you don't see email links below, enable "javascript" in your browser.

mail to

mail to

Humming bird page

More information on hydropower at the microhydro webportal

Last modified on 03 May 2003  since May 2003
FastCounter by bCentral