In the News:

May 7th, 2000:
My human power research outlined on page 2 of Sunday's Kathmandu Post

Ghatta Charging Station - Human Powered Generators - White LEDs
WireBridge - WireBike - MicroHydro Battery Charging - PicoHydro
Contacts / Sources

Last Update: May 16, 2000


I have conducted research on low cost solutions for domestic power generation in Nepal under the auspices of a Fulbright scholarship. This research looks at Hydro, Human and Solar energy sources for low power lighting systems: specifically using white LED technology. On this webpage I am faced with the challenge of filtering this information down to only the highlights of my research and product development in each of the areas I have worked.

Ghatta Charging Station


Traditional Nepali Ghatta


Much of the countryside in the middle hills of Nepal is speckled with small watermills (ghattas) used for agro-processing. These ghattas harness the power of diverted water and a 1-2 meter head to turn circular (~75 cm diameter) stone grinders at about 60 rpm. This rotational energy is only utilized during the daytime and has great potential for small scale rural electrification.


Ghatta in Action

One simple approach would be to install bicycle dynamos on the grinding stone. When the grinder is running it would be also generating electricity stored in 12V batteries. At night a smooth plastic sheet could be put in between the stones to keep the grinder from wearing out and decrease the friction of the system. Installing 10 dynamos around the wheel would generate about 100W and cost about 3000 Rs. Individuals would be able to purchase 12V batteries based on their lighting needs and perhaps pay a charging fee to the mill owner. Such a system would be quite applicable for white LED lighting.
However, in initial testing, bike dynamos have gotten quite hot after an hour of use. Because these dynamos are designed to be cooled by fast moving air, a heat sink may needed when they are used in stationary operation. Other alternatives include hooking a 12V/40A jeep dynamo (3500 Rs) on the ghatta axle with a bicycle tube as the belt drive. Motorcycle alternators have been looked at as well, although they are currently built inside the motor casings. New car alternators are the best candidates for the ghatta generator. The Nepal Ghatta Project was formed to create a charging station using this solution.

Human-Powered Generators


Nepal Light Project's Pedal Power Generator


The majority of the unelectrified population in Nepal do not have access to a permanent water source from which they could generate electricity. However, human power is universally available and is quite applicable if only small wattage is required. Several groups have come up with schemes for incorporating pedal generators in rural electrification projects for remote areas on Nepal. One such group is the Nepal Light Project led by Professor Irvine-Halliday of the University of Calgary and the engineering students at the Center for Energy Studies in the Institute of Engineering here in Kathmandu. Professor Irvine-Halliday's generator is designed specifically for White LEDs and battery charging. Using the system he designed, 20 minutes of extremely minimal pedaling gives 3-4 hours of light to a household (using 6 White LEDS).


A 350 Rs Bicycle Dynamo available in Kathmandu


I have started putting together my own prototype using locally available parts. Bicycle dynamos imported from India and China are available in Kathmandu for 350 Rs. (US$5.00) Connected to a multimeter I was able to verify that they put out about 12V AC and 6 Watts of power. Installing 10 in parallel on a used 12-speed bicycle should generate over 50W - enough to power a B&W TV as well as adequate White LED lighting.
Human-powered generators will be a major focus of my research for the next months in Nepal. However, as a colleague pointed out, much of the potential users of these systems (especially the women) are already overburdened with significant manual labor. A more elegant, although location specific, solution would be to leverage the work already exerted in existing daily labor. An example could be a simple dynamo hooked onto a hand-grinder. Part of my upcoming fieldwork will be to look for these applicable daily laborious chores.

White Light Emitting Diodes (LEDs)


The "Kerosene Killer"!


A White LED table lamp, emitting enough light to read comfortably, will cost about $6.00 and be powered by less than a third of a Watt. With 3.2 million houses in Nepal currently being exposed to the health hazards of kerosene lamps, this new technology (which was not an option even a year ago because of cost) can have a widespread impact almost immediately. These "Kerosene Killers" have an extremely long lifetime and are virtually indestructible (no filaments!). When bought in bulk, each white LED costs little over US$1.00.

Technical Specifications for the 5.6 Cd White LED


The 8 LED Ceiling Lamp


Price: $1.00 - $3.00 (and dropping)
Voltage: 3.6V
Wattage: .072 Watts

Results from Initial Field Tests

Prototype lights were loaned to families and young students from the Kavre and Tanahu districts. Many of the people surveyed were willing to buy the lights above cost.



The 3-LED Table Lamp


Accord to a recent report by the Center for Energy Studies, powered by3 rechargeable D-cell batteries, the Table Lamp gives over 200 Lux @ 30 cm for over 17 hours of use. The voltage of the batteries during these 17 hours drops only .33 V, so a very small power source is necessary for weekly recharging. Unlike other power generation schemes, solar power can scale with the energy required. A 1 Wp solar panel generates more than enough power for 2 Table Lamps and a Ceiling Lamp. (assuming a 4 hr/day charge time) The cost breakdown of this potential solar lighting system is as follows:
• 1 Solar Module 6 V, 1 Wp - $10.00
• 6 Rechargeable 1.2V D Cells / 6 Ah/6V cells: $17.00
• 2 Table Lamp (3 LEDs/lamp): $12.00
• 1 Ceiling Lamp (8 LEDs/lamp): $16.00
• Cables/Fixtures/Accessories: $10.00
TOTAL: $65.00
Since such a low wattage is required (about 1 W/household), this new technology also opens the door for many inexpensive, small generators such as the 6 W, 350 Rs bicycle dynamo.

Ramifications of White LEDS within current MHP Projects



A 3-LED Flashlight



Attenuation is a major reason why many villages nearby MHP sites are not currently powered. Running 220AC over 275 m using a 3/22 cable drops the power down to 196 AC. However, if then converted to DC, it gives 3.9V which can be distributed within a community. Assuming that every household served by the MHP gets the standard load of 60W, this entire community of 30 households could be electrified with the power allocated to one.



First Ride Across the Sun Kosi!


The first four wire bridge during its inauguration


A WireBridge is a cost effective river crossing solution for communities who do not have the means to build the more expensive suspension bridges. It is comprised of either two or four wires (which can span 80/160 meters respectively) and a set of trolleys with an attached carriage. People control the position of the carriage with the rope connected to either side. To cross the river, a Nepali would pull the carriage towards her and boards it when it is at an appropriate distance. Letting go of the rope, she is then propelled more than halfway across the river with just the potential energy from the slack in the wire. Gentle tugs on the rope will get the carriage the rest of the way across.


The redesigned trolley


My first task was the redesign of retainer for the wheels. The trolley wheels were not running smoothly because they needed more freedom for lateral movement along the wire. After a few iterations on various designs we have currently settled on putting the trolley retainer outside the wheelbox, giving the wheels much more lateral play. I drew up the engineering drawings for this design and worked with a skilled local machine shop who was able to manufacture the wheelboxs to spec on their second try. This final design is still being tested, but the results look promising.


Helping set up the carriage


I also got the opportunity to lend a hand to the construction project of EcoSystems' first four-wire bridge. We floated most of our supplies down the Sun Kosi river and when we beached at the site, it was a matter of erecting the posts and stringing the wire. Sounds simple enough, especially with the 20+ volunteers from the village, but it took a solid three days of work.


The WireBike is part of a grander scheme called the WireRoad. The idea is simple: traditional roads have had a devastating environmental impact on Nepal - yet they continue to be built at an even faster rate simply because there is not other viable alternative for transportation. The WireRoad is a high tension cable that can be used to transport goods and people from one village to another without bringing in the bulldozers. We are envisioning two types of locomotives: pedal-powered and electric. I have spend the majority of my first 90 days here working on a pedal-powered design that can easy be retrofitted to take an electric motor on the drive shaft: the WireBike.
Many different designs have been pursued, and I have had some great help from people abroad over email as well as here in Nepal. When I arrived, I already had a prototype vehicle designed at MIT to assemble and test. Although there were some inherent problems with the prototype it was a great first stab at a problem more complicated than it initially appears.


Design 1 - MIT prototype


The problem lies in the necessity of freewheeling reversibility: the bike needs to be pedaled back and forth along the same wire, although constraints intrinsic to the WireRoad make it impossible for the wheels to be simply lifted off the wire and flipped around. The initial prototype solved the reversibility issue by creating a symmetrical upper frame attached to the wheels. The recumbent seat and bike frame would detach from this structure and be flipped around for travel in the reverse direction. Because of traction issues suspected going uphill, the vehicle had both front and back wheels powered by a single very long chain. Many problems were solved when we modified the vehicle to be only front wheel drive - although alignment issues with the chain still occurred frequently and the large "toothed" bike rim wouldn't always catch the chain.

Reversible Hub

Reversible Hub


I began design work on a new type of system with the sole focus on a simple and robust solution for reversibility. I did not want the user to have disassemble anything - switching the freewheeling direction should require only one person and should take seconds, not minutes. With this priority in mind, gears and two-wheel drive were sacrificed for the sake of robust simplicity. The initial design involves two chairs centered and facing each other with a crank shaft connected to a special "reversible" bike hub. In order to change the freewheeling direction quickly and easily, the mechanical principle of a socket wrench must be incorporated into this hub.


Design 2 - Easy Reversibility


This was achieved with the help of Mr. Sharma and the lathe operators in the machine shop at Balaju Yantra Shala. We took a standard rear bicycle hub and cut it in half - taking out the bearings. We then welded another section onto the hub essentially elongating the threaded portion. With the elongated threads and the bearings reinstalled inside the new section, two opposing bike sprockets were screwed on to have different freewheeling directions. A much larger motorcycle gear is then added to the original side of the hub and connects with the drive wheels from above. To reverse direction the driver simply moves the chain less than an inch from one sprocket to the other and sits down in the opposite chair.
Just recently, a very simple solution came from the plantations in South America where a similar wireroad has been constructed to transport bananas from the field. This solution, although the reversibility and gearing is unclear, seems the most straight forward for manufacturablity. It uses standard parts and leverages most of the aspects of the common 18 speed bicycle with the rear sprockets offset to be directly above the main drive system.
At the moment these are the three designs on the table:
• "Design 1" leverages the structure of the initial prototype but uses much more robust wheels and chain. Its main drawback is the complexity involved to reverse directions.
•"Design 2" is a new design using the custom reversible hub. This creates an easy and fast way to change directions - but is untested and parts of the frame still need more design work.
•"Design 3" is the easiest to build and test - however it is uncertain how to make it reliably reversible and geared appropriately.



A Working Prototype of the WireBike

Which design to pick depends on how quickly a working demo is needed. To get the quickest working prototype "Design 3" seems appropriate, however it is questionable if that design can be the long term solution for WireRoad transportation. "Design 1" would be the next easiest solution simply because most of the engineering has already been accomplished and only gears/wheels and chain need to be purchased. However, the design should not be pursued if its method of reversibility is unacceptable for operators or conditions in Nepal. The "reversible hub" solution of "Design 2" seems to be the most robust yet will take the longest to construct.

Micro-Hydro Battery Charging

Lighting demands in rural areas are characterized by high peak loads in the morning and evening yet the energy supply from a micro-hydro plant is typically flat. Micro-hydro plants (MHP) in Nepal have been designed to meet these peak loads and therefore generate much wasted off-peak energy. Battery charging is an obvious, yet underutilized way to leverage this off-peak energy.

battery charger

DCS Battery Charger


DCS Study:
In1997, Development Consultant Services (DCS) began two pilot projects aimed at using a micro-hydro plant's off peak energy to charge up batteries for unconnected neighboring households within a few hours walk from the plant.
Equipment: (US$1=68 Rs)
• Because of their efficiency and cost, 20W fluorescent lamps with DC chokes were used. (350 Rs) (The 12V rated vehicle headlight lamps were not selected because of they consume more energy and emit less light compared with fluorescents.)
• 70AH 12V Solar batteries, although significantly more expensive than motor vehicle batteries, were selected because of their allowable depth of discharge and life span. (6440 Rs, 24 kg with electrolyte, 5 year life span, 80% depth of discharge) (Motor vehicle batteries begin to fail after less than 100 cycles of discharge to 50% capacity)
• A 70-150W battery charger was designed and built to recharge a 70Ah battery within 12-14 hours from the off peak (daytime) electricity of a microhydro plant. (5745 Rs)
Although most appreciated the lighting project, not all the users wanted to keep the battery lighting system for several reasons:
• Acid Spillage: although they were wrapped in a polythene sheet, the batteries only had push-on caps which contained charged acid with a density of 1.280.
• Weight: some villagers couldn't carry the batteries themselves (24 kg) and had to hire porters.
• Inconvenience: because of the 12-14 hour recharge time and waits to use the charger, people needed two batteries to make sure they have power every night.
• Sustainability: The 5 year life span of the 5745 Rs batteries gives a cost of over 1000 Rs/year for batteries alone. There is no way to guarantee the donor agency will continue to supply/subsidize batteries.  
LEDCO Study:
LEDCO gets around the danger of carrying lead-acid batteries with a design called, 'Hybrid Micro-Hydro'. One proposal (awaiting funding) incorporates both AC lights run directly off the MHP as well as several "battery banks" powering the DC florescents mentioned above. This allows the micro-hydro plant to be sized for the constant power demands of the battery chargers instead of peak load. During off-peak hours, the MHP continually recharges (tops off) the bank in preparation for the peak loading times. Theoretically, this enables a 3KW Hybrid MHP to power more households than a 10KW traditional MHP - cutting the cost of the MHP considerably. However, the critics point to the fact the cost battery banks offsets this savings significantly (if not completely). However, when made aware the technical specs of new White LEDs, they are now in the process of creating another proposal using this new technology with only a fraction of the battery cost.  


laos pico

Pico-Hydro in Laos


In Vietnam, China, and Laos, there are pico-hydro (100-500W) generators that have significant potential in Nepal. The facts below are taken from a summary of a report by IDE:
• Standing units (draft tube required) cost between US$15-18.
• Sitting units (pipe connection required) cost between US$30-35.
• They require significant flow rate (30 liters/sec) but low head (1-2 m).
• There is an estimated 120,000 installed in Vietnam.
• Most of those installed are not working with optimal efficiency.
• Some are electrically unsafe. (6 known deaths due to exposed wires, etc)
• Theft of these units is a problem.
• The bearings, windings (generator coils) and seals fail on a regular basis, but can usually be repaired by locals cheaply. New bearings cost US$1.10 and new windings cost US$1.65.
• The poor quality is seen only as a hassle rather than an impediment to purchasing.
• Installation is done by the purchaser.
• Operational problems - wet season: debris during flooding causes intake strucures to be washed away/generators get submerged. dry season: not enough flow.
• Large number of manufacturers but no real distribution network.
• Main competition is "waiting for the grid"

Technical Specifications of the 300W Design


Solid Model of Units at NHE


• 6 pole single-phase 230 Vac machines (60 Hz @ 1200 rpm)
• Propeller turbine
• Very peaky efficiency curve with max power produced for a very narrow range of head/flow - - expected output for design purposes should be 33-50% of rated output. (halving the flow rate drops the output by over 66%)
• Load/frequency controller may be needed for electronics you don’t want to fry.

Key Characteristics of a Potentially Successful Market in Nepal

pico in laos

Getting the required 1.2m head


Part of my upcoming fieldwork will be to look for these types of characteristics:
• River with steep grade so that an easy diversion structure can be made.
• Irrigation projects where the civil work has already been complete and a 1.2m drop is available.
• Consistent water supply (more than 30 l/s during monsoon, 15 l/s during dry season)
• Easy access to the supporting infrastructure (providing spare parts and repair) is critical
• Strong community groups: farming co-ops, schools, women's groups
• Not about to be grid connected
• Some people with technical/mechanical skills in community
• Use money in trading
• Have credit experience
• Significant portion of the community has sufficient income



Pico-Hydro Lab at Nepal Hydro and Electric (NHE)


To bring this product to Nepal, the cost will more than double:
• Vietnamese Standing Pico Hydro System -- $15.00
• Shipping -- $15.00
• Import duty (10%) -- $3.00
SUBTOTAL: $33.00
If a load controller is required, this price will jump to at least $60.00. It may be possible for a good battery and rectifier to take the place of a load controller - but this needs further investigation.
Because the cheapest peltic set in Nepal (200W) costs over 20,000 RS (~US$300), even at a price point of $100.00 this represents a significant opportunity.

other pico idea


This is a pico-hydro at DCS requiring very high head but low flow. The runner is housed within the pipe itself.





Gyanendra Bahadur Bhandari
Section Chief (loans for micro/pico hydro)
Agriculture Development Bank (ADB)
Work: 292391, 292360 ex:207
Stewart Craine
Jeff Dickinson
Dave Irvine-Halliday
E.E. Professor, University of Calgary Light Project
Dilliraman Neupane
Alternative Energy Promotion Center (AEPC)
Work: 522520
Bikash Pandey
WinRock, REPSO Nepal
Work: 253687
Ghanashyam Ranjitkar
Electrical Engineer
Intermediate Technology Development Group (ITDG)
Work: 529815
Kamal Rijal
Renewable Energy Specialist
Work: 525313
Jagan Nath Shrestha
EE Professor, Director, Center for Energy Studies
Institute of Engineering
Tribhuvan University
Work: 532235
Dr. Krishna Shrestha
Tribhuvan University


Design, Construction and Field Testing of a Low Cost Lighting System for the Rural Nepal. (White LEDs) for: REDP by: Center for Energy Studies, Institute of Engineering, Tribhuvan University, April 2000.
Redesigning rural electrification in developing countries, for: LEDCO by: Stewart Craine.
Micro-Hydro Charged Rural Battery Lighting Project, for: DCS by: Ghanashyam Ranjithkar, January 1999.
Final Report of MHP Operator Training Evaluation, for: ITDG by: Kayba Niroula, IEDI, June 1999.
Micro-Hydro Support Programme Preparatory Works. for: Dandia ESAP by: ITDG, November 1999.
Pico-Hydro for Village Electrification in Nepal. for: DID by: P. Maher, November 1998.
The Techno-economic Performance of Water Turbines in Rural Communities of Nepal. for: ITDG by: New ERA, August 1993.
Rijal, Kamal (editor), Energy Use in Mountain Areas, ICIMOD, February 1999.
Manual for Survey and Layout Design of Private Micro-hydropower Plants, ICIMOD, 1999.
Installation and Commissioning Manual for Pricate Micro-hydropower Plants, ICIMOD, 1999.
Operation and Management Manual for Private Micro-hydropower Plants, ICIMOD, 1999.
Maintenance and Repair Manual for Private Micro-hydropower Plants, ICIMOD, 1999.
Rijal, Kamal (editor), Renewable Energy Technologies, ICIMOD, May 1998.
Manual of Rural Technology with Implications for Mountain Tourism. ICIMOD and CRT, December 1997.
Financial Guidelines for Micro-hydro Projects. ITDG, 1997.
Micro Hydro End Uses Catalogue. ITDG, 1997.
Micro Hydro, Year Book of Nepal (July '98-July '99). CADEC, November 1999.
Management Guidelines for Isolated Micro-hydro Plants in Nepal. ITDG, 1999.
PicoHydro, issue 4, April 1999.
Renewable Energy Technologies, ICIMOD Newsletter, no. 30, summer 1998.
PicoHydro, issue 3, October 1998.