Hydroelectric Resource Evaluation:  Cloudbridge and Gode Homes

Termometro, San Gerardo de Rivas, Mount Chirripó, Costa Rica

For:  Ian Giddy and Tom Gode

By:  Paul Collar

Osa Water Works

February 13, 2006

Overview

Osa Water Works was contracted to undertake a feasibility analysis for the installation of a hydroelectric power generation system to supply the needs of two semi-adjacent residences, both located on the flanks of Chirripó Mountain, which translates from the native language as “Mother of Water” Mountain.  One of the homes (Cloudbridge) is an existing residence already powered with 12-V solar power, the second a recently finished home. 

Domestic Power Requirements

Table 1 presents the anticipated power demands for the individual homes, based on anticipated infrastructure and projected usage patterns.

Table 1.  Anticipated Power Demand Calculations

On the basis of the components listed, each house is expected to have a power demand, therefore of around 6 kw per day, for a total of 12 kw for both facilities.  To achieve the power demand required for the two facilties requires the ability to generate, therefore, 500 watts per hour, and to have the battery capacity to absorb and disburse this charge. 

Hydroelectric Potential

The stream to be used to provide the hydraulic energy to generate electricity is one contained on the property in question.  A conceputal diagram of the optimal stream intake point and other structures relevant to the discussion that follows is shown in Figure 1.  As shown the elevation difference between the Cloudbridge home and the water tank is somewhat less than the elevation drop between the water source and the tank.  In practice, a final generation point can be placed below the elevation of the existing home in order to obtain comparable hydroelectric potentials for both.  However, for the moment, the Cloudbridge home elevation was used for the ensuing analysis.

Figure 1.  Approximate distances between property features and the approximate elevation of features normalized to the elevation of the Cloudbridge home.

At the time of the field survey, a water flow in the target stream suggested that diversion of 50 gpm would be entirely reasonable, which corresponded to less than half of the approximated water flow at the proposed collection point at the time of the survey.  This is not an excessive withdrawal amount, especially considering the downstream gains the stream achieves from springs located below the target extraction point.  Finally, the water is to be returned in its entirety to the watershed upon discharge from the hydro turbine. 

Hydroelectric potential is calculated from formula (1)

                                     P = Head * Flow * 0.18 * Turbine Efficiency                 (1)

Where P is the hydroelectric poential in watts / hour, Head is the vertical drop measured in feet, Flow is the flow rate measured in gallons per minute, 0.18 is a units conversion coefficient, and Turbine Efficiency is assumed to be 50% (reasonable for micro-hydro applications.  Applying this formula to the different possible system configurations and a flow rate of 50 gpm, the hydroelectric potentials calculate to the following:

Intake – Water Tank              :  810 watts per hour

Water Tank – Cloudbridge    :  540 watts per hour

Intake –Cloudbridge              :  1.35 kw per hour

Even the very lowest energy yield among these configurations, the hypothetical run from the water tank to the Cloudbridge house, 540 watts per hour, exceeds the anticipated power demands for both facilities.

Alternative Configurations.

The great disparity between hydroelectric potential and electrical demand for the two facilities suggests that barring anticipated expansion of the facility or the desire to use supplemental energy for purposes not contemplated at the time of the survey, that the electrical demands of both facilities can be achieved with a single pipeline and generation point at the water tank.  However, to do this, in effect requires that the entire power system, including one large inverter and the battery bank be located at the water tank site.  This presupposes the construction of a secure power bodega on top of the tank, which introduces construction costs that would be considerably higher than for two separate independent systems.  Moreover the transmission distances involved would require the use of either very large gage cable or the installation of a custom transformer to boost the voltage to 240 to be split into two transmission lines of 120 each to service each of the houses.

Even with the installation of two independent systems, i.e. paralllel or back-to-back in-series penstocks with two separate turbines and separate inverters and battery banks, the distance from the water tank to the Gode house is still sufficiently large that either prohibitively large cable would be required to convey raw power to an energy center in that household, or the Gode power center would have to be located on top of the water tank and the power distribution to the Gode house achieved through smaller cable. 

Therefore, even with split systems, we cannot get away from the expense of construction of a power bodega at the site of the existing water tank.  In effect, the challenge in the most appropriate configuration is the actual location of the water tank, which is inconveniently distant from both houses.  While the pelton wheel does produce alternating current (which has lower line losses than direct current), the raw power generated by the turbine is not regulated and of a variable frequency that greatly increases its line losses.  The inverter in effect regulates and polishes this output into an alternating current that has much lower lines losses, enabling the deployment of much smaller and considerably less expensive cable

It would appear, therefore, that there are three alternatives available worthy of consideration, but in two of these alternatives the existing water tank is not even used.  These three alternatives are as follows:

Option 1:  Central power system at site of existing tank.

A conceptual diagram is shown in Figure 2. 

Figure 2.  Conceptual diagram of a centralized power system located at the existing tank site. 

This system presupposes a 6000 watt 12-Volt inverter, a battery bank of 8 12-volt batteries, a 120-240 transformer and buried #2 electrical transmission lines carrying ready-to-use AC power to each of the residences.  In this configuration, each house would have a breaker box and internal electrical would be for all practical purposes the same as connecting to grid power.  While this is a reasonable system, it has a total hydroelectric potential of only 19.4 kilowatts per day at 50 gpm.  In practice two-inch pipeline can carry considerably more water than 50 gpm, and the stream has more than this flowrate available.  However energy in excess of the rated system capacity would have to be dissipated at the tank site instead of utilizing productively for heating water or operating 12 volt utilities (like refrigerators), etc.  One major consequence of a power center located at the tank site is that 12-V refrigeration could not be used at either of the houses, which would raise the daily energy demand estimate of each house to 13 kw and the collective demand to 26 kw, an energy demand not satisfied by the resources at the water diversion rates considered in the calculations.  Therefore, before this alternative may be considered additionally, it would be necessary to ensure that 70 gpm could be captured at the intake, which would provide the energy necessary for two medium sized refrigerators at both households, and to possibly increase the invertor capacity to around 8000 watts.

Option 2.  Central power system at the Cloudbridge homesite.

A conceptual diagram of this alternative is shown in Figure 3.

Figure 3.  Conceptual diagram of a centralized power system located at the Cloudbridge home.

This would essentially be an exact duplication of the Option 1 configuration described above, except it would be located adjacent to the Cloudbridge home.  If the entire head of 300 feet is used and presuming a conservative 50 gpm water flow, this power corresponds to 32.4 kw per day, which is 30% more power for approximately similar costs as the first alternative.  This system would allow for heating of water with excess power, perhaps even to make it hot, depending on actual electrical usage patterns.  One big advantage of this configuration is that the ownership is not locked into the intake location we visited.  Since the ownership does not need 32.4 kw per day, a lower intake location can be identified that will give what is needed with an adequate margin for safety and expansion, which will save on pipeline costs.  Moreover, since there is more water in the stream the farther we move downstream, it will be possible to increase the pipeline diameter to 2.5” or even 3” diameter pipe to achieve comparable power with lower head and more water.  This would require a second field determination to determine what the practical alternatives are and how the economics balance between pipe costs for varying diameters and run distances.  Under Option 2, the Gode household would require AC refrigeration, while the Cloudbridge facility could deploy a 12-Volt refrigerator.  The additional power sink of the AC fridge at the Gode residence would be easily accommodated by the capacity of this configuration.

Option 3.  Separate systems in Parallel

A conceptual diagram of two parallel systems to service each residence is shown in Figure 4.

It is entirely reasonable to construct two completely independent system from scratch.  In this case, the water tank would be left alone as the Cloudbridge home’s water supply tank, and a new pipeline from high in the watershed would deliver water to near the Gode household.  In order to re-use this water, a second pipeline could then convey this water to the Cloudbridge home for power generation there (a configuration that I have coined “separate systems in series.”  However, owing to the greater versatility offered the Cloudbridge home, I would consider discharging the water used for the Gode residence into the environment and simply run a second “parallel” pipeline from a lower part of the river to service the Cloudbridge Home.  This parallel configuration would require two separate pipelines, two intakes, two 2000-watt, 12-Volt inverters, and 8 12-Volt batteries and associated charge controllers.

Figure 4.  Parallel and separate micro-hydro systems.  One variant of this configuration is to place the intake of the Cloudbridge penstock at the Gode point of generation to use the same water in separate systems in series rather than parallel.

Economic Analysis

            There are a wide range of makes and models of equipment and batteries with different pricing structures, and the distances used for of pipe and electrical cable estimation are only estimates.  The costs that I have used in this analysis are from a retail cost sheet from my San Jose distributor and estimates in some cases, such as for the step-up transformer, which is a custom piece of equipment that must be fabricated according to duty-specific criteria.  Additional economic considerations include the costs of 12-Volt refrigerators, which are around $1000 more expensive than off-the-shelf AC fridges.  Also, at the end of the day, it will be fruitful to compare the capital costs of step-up/step-down transformer inclusion for Option 2 to reduce line losses and permit less expensive smaller-gage transmission cable.  Finally, the estimates do not include cost estimates for physical structures such as a power bodega on top of the water tank for Option 2.  The numbers presented in Table 2, therefore, are not intended to be firm prices, and this by no means should be considered a bid.  The analysis is simply a tool to appreciate the cost differences between the three alternatives listed.

Table 2.  Cost estimates for the major components and installation costs for the three options described in the body of this report.

As Table 2 reveals there is essentially no price difference between any of the three configurations, and since Options 2 and 3 provide substantially greater power, this would appear to argue strongly that Option 1 be discounted.

However, costs have not been approximated for the power bodega that would be required to house centralized systems in Options 1 and 2, and this is likely to be an additional $10,000 or so.  Moreover, Option 3 in this case presumes wholly separate systems and does not account for the option of reducing the length of the Cloudbridge penstock, which will carry supplemental savings.  It is presumed that the capital costs of adapting existing structures or storage facilities to accommodate individual battery banks and electronics for the parallel systems will not be a capital intensive element, though this will depend on ownership preferences.

In my opinion, provided the existing bodega on the Cloudbridge home site can be adapted to serve as the central power bodega, it is an economic and functional toss-up between centralized power generation at the Cloubridge home and separate parallel systems.  However, the option of either parallel separate systems or back to back systems in series has the advantage of each household being wholly independent of the other.  While separate systems in parallel offer the Cloudbridge home considerable leeway in refining a system to conform to actual needs versus overdesign to provide for water heating, separate systems in series do not offer this versatility since the intake point would be fixed at the point of generation from the Gode household.  While additional capacity can be obtained by deploying the Cloudbridge point of generation lower on the slope, this option requires careful consideration of corresponding line losses and the additional costs in cable to overcome these, so there is a point of diminishing returns the further the point of generation is moved from the inverters and battery banks.

Conclusions

The property on the slopes of Costa Rica’s tallest mountain, Mount Chirripo, possesses abundant hydroelectric resources greatly in excess of the power demand anticipated by the two households.  However, the stream is not large enough for AC Direct power generation, capable of satisfying the facility demands.  Hence, micro-hydroelectric power generation is the only alternative available to tap the resources.  However, high line losses of raw hydropower require for the sake of efficiency that inverter(s) be located close to the point(s) of generation.  This requires either a central power station that supplies both homes with AC power, or separate systems.  While this central power station can be erected at the site of the existing water tank, this would produce about 30% less power than deployed at the Cloudbridge house site and carry the same costs.  Barring aesthetic considerations, the two most reasonable alternatives would be centralized power at the Cloudbridge house site OR separate and parallel systems.