Hydroelectric Resource Evaluation

Finca [----], Palmichal de Puriscal, Costa Rica

For:  [-----redacted------]

By:  Paul Collar

Osa Water Works

June 11, 2008

EXECUTIVE SUMMARY

The 247-hectare Finca Adiabet has at least two miles of Río Negro river frontage spanning a vertical relief of around 2000 feet.  The river has a flow rate that in conjunction with the slope of the terrain makes an ideal location for a high-head low-flow AC-Direct application.  A rough preliminary economic analysis of a 130-kilowatt configuration suggests a project payback period of 5.56 years and potential revenues of $85,410 per year thereafter, predicated on power sales at 75% of domestic power rates presently charged by the power company.  The entire feasibility of a hydroelectric installation depends upon two factors independent of the physical property resources that must be negotiated and contractually assured by the Republic of Costa Rica in order for the project to be viable.  These are:

·         Negotiated terms for long term sales of power to the national electrical grid, and

·         Water concession that permits enough flow rate to make the project economical.

The overall project viability will be constrained and limited by the flow rate during the dry season.  In order to ensure that final decisions are reached during the 2009 dry season, it is necessary to begin the application process with ICE and SETENA as soon as possible.

INTRODUCTION

Osa Water Works was hired to conduct a hydroelectric resource evaluation for the 247-hectare Finca Adiabet, located in the mountains of Acosta Cantón, near the town of Palmichal in central Costa Rica.  The objectives were to determine the hydroelectric potential of the Río Negro along the river reach that bounds the property, and to undertake a preliminary feasibility evaluation for a hydroelectric facility for sale of electricity to the national power grid and to satisfy the power demands of the property.  This report summarizes the study and findings.

LOCATION

Figure 1 shows the property boundaries of Finca Adiabet, which actually includes several distinct properties that were assembled to comprise the new 247-hectare tract of land.   Figure 2 shows the boundaries on a aerial photograph.

Figure 1.  Plano of Finca Adiabet

METHODS

Due to the large size of the property in question, limitations on time, and the difficulty of access, it was possible to explore only a portion of the watershed on the day of the field survey.  A Garmin GPSmap 76Cx Global positioning system was used to log waypoints examined along the survey transits.  A Casio digital altimeter was used to determine elevations.  Flow rates were estimated based on observation.

Figure 2.  Aerial photograph of Río Negro watershed, showing the boundaries of Finca Adiabet.

FIELD SURVEY RESULTS

Figure 3 shows the GPS track of the vehicular and pedestrian transit undertaken in the field superimposed on the site map from Figure 1, the aerial photograph from Figure 2, and on an aerial photograph obtained from Google Earth.  Appendix 1 shows the raw field information collected for each of the waypoints visited.

The flow rate of the river was estimated to be several thousand gallons per minute at all the portions of the river visited.  Verbal estimates of dry season flow suggest that the flow rate is perhaps as little as one third of that observed on the field survey for this report.  Even at reduced dry season flow rates, it is apparent that the Río Negro is likely to sustain an economically viable hydroelectric power generation capacity.  Still, project limitations will ultimately be predicated not on the flows observed during the field survey for this report, but on water diversion permit allowances made by regulators, which will be based on dry season flow rates.  Correspondingly, final engineering and project sizing will necessarily depend on a dry-season evaluation of flow rates and permissible diversion rates.  For the purposes of this resource evaluation, estimations of a ceiling on reasonable water extraction rates are assumed to be 3000 gallons per minute (gpm) at the highest point of the watershed reached during the site survey.  Flow rates are expected to decrease moving up the watershed, though this would need to be corroborated by a site visit to determine the extent to which the entire watershed may present viable options for a variety of alternative hydroelectric sites.

Figure 3.  GPS Track Overlays on:  site plan, aerial photo, and Google Earth photo.

DISCUSSION

Figure 4 is a slope diagram of data points taken along the road located on the farm on the other side of the Negro River from Finca Adiabet.  Although the road diverges from the elevation of the river in places, the highest point on the graph and the lowest point on the graph were measured at the river level itself, and the graph is a reasonable approximation of the stream gradient profile.  As Figure 4 reveals, there is no appreciable break in slope along the approximately one mile length of river that was transited from the base of the property until the end of the gravel road.  While it could be argued that the first 1000 feet and the river reach from 4000 feet to 5500 feet have higher slopes than the reach of river between 1200 feet and 4000 feet, the difference in slope is not likely to give rise to significant economic favorability for hydroelectric power generation along short stretches of the river.

Figure 4.  Stream Gradient Profile.

HYDROELECTRIC POTENTIAL

Hydroelectric potential is approximated using the empirical formula given as Formula (1).

P =  Q  x  H  x  0.18  x  e                              (1)

where P = power in watts, Q = flow rate in gallons per minute, H = vertical relief (or head) measured in feet, 0.18 is a units conversion constant, and e is the efficiency of the turbine.  For micro-hydroelectric turbines, an efficiency of 50% can be assumed.  For mini-hydroelectric applications the efficiency is higher, typically around 65%.

Applying equation (1) to the reach of river transited, the head of 670 feet, a flow rate of 2500 gpm, and a turbine efficiency of 65% corresponds to a hydroelectric potential of 196 kilowatts. 

Since the headwaters of Río Negro are reportedly perennial at an elevation of 1300 vertical feet higher than the highest point visited during the field survey for this report, it is evident that Finca Adiabet has the resources to support a variety of hydroelectric configurations including the possibility of multiple installations in different reaches of the river.

DESIGN AND ENGINEERING

In general, AC-Direct mini-hydroelectric power generation plants are either high-head, low flow, or high flow, low head applications.  Traditional hydroelectric plants at the base of reservoirs are examples of the latter, employing gigantic penstocks across vertical relief of commonly less than 100 feet.  Because dry season flow rates on the Río Negro are reported to be modest, the most likely hydroelectric configuration for a year round hydro plant is expected to be a variant of a high-head, low-flow application.  This will take the form of a relatively long modest-diameter pipeline, in which hydroelectric power generation is optimized for head preferentially over flow rate.  As an example, let us assume for the sake of argument that an eight-inch pipeline diverts water from the headwaters of the Negro River at the top of the farm into a penstock leading to the base of the property.  Although slopes are expected and reported to be steeper in the headwaters than in the portion of the river visited and documented in Figure 4, let us assume for the purposes of this analysis that the slope is the same as that shown in Figure 4.  The following design variables therefore apply to our hypothetical hydroelectric configuration.

Head:  1400 feet

Flow:  750 gpm

Distance:  2 miles

Head Loss:   125 feet

Net Head:  1275 feet

Power:  112 kilowatt

On the basis of the foregoing calculations it is clear that a relatively robust mini-hydroelectric application is feasible on the Río Negro.  Whereas the water flow rate accommodated by an 8” pipeline was used for the calculations, a much larger pipeline diameter could be used without drying up the river, certainly within the reach of river visited for this report.  Part of the iterative design process in a case in which there are a number of choices for the location of intake and generation is to determine if it is more economical to have a lower diameter, longer pipeline that capitalizes on greater head, or a shorter, larger diameter pipeline that sacrifices some head to gain additional flow.  All things being equal, head is usually favored over flow rate where both are in play.  In practical terms, the specifications and pricing of materials of construction carry important design limitations in this evaluation.  First consider the penstock.

Among the only two plastics that are potentially viable as penstock pipe,  poly-vinyl chloride (PVC) is reportedly 15-30% less expensive than high-density polyethylene (HDPE).  Both materials have standard pressure ratings topping out at around 250 pounds per square inch (psi).  Across the entire stated head range of 1400 feet in our hypothetical example, a total of 600 psi of pressure is developed by the density of standing water.  Clearly, in a hypothetical penstock running the length of the property, steel pipe would be required to withstand the static and operating pressures developed.  While PVC is manufactured in Costa Rica by two different companies, steel pipe must be imported.

At $23.50 per foot, a price quoted for 12” Grade B steel with ¼” wall thickness capable of sustaining pressures up to 1200 psi, the pipeline described above would run $250,000 for the pipe alone FOB USA..  The same pipe in PVC has a retail price of $46.70 per foot in Costa Rica.  Once transportation, import tariffs, and price inflation are factored in, it would appear that the pricing of plastics in Costa Rica is somewhat more expensive to the pricing of steel imported from the United States though considered comparable for the purposes of this report, a parity that was corroborated by obtaining a quote for steel pipe from a single Costa Rican steel importing house.  While it will be necessary to source alternatives for the best pricing in a final engineering and design phase, for the purposes of preliminary feasibility, the pricing of domestic plastics is assumed on the basis of preliminary costing to be approximately the same as the pricing of imported steel pipeline.

Having tentatively assumed for the purposes of this analysis that there is little economic advantage among different pipeline materials of construction, the remaining design criteria most prominently include penstock length, diameter, and the net head.  The table below summarizes ballpark capital costs for a 1-kilometer length of penstock assuming the river slope gradient shown in Figure 4.  The analysis assumes pipe ranging in size from 8” to 15”, the latter representing the largest suspected possible flow rate that the river is expected to carry without adverse  environmental or aesthetic consequence.  Flow rates were estimated on the basis of a velocity of 5 feet per second, a velocity that will stay within the regime of flow efficiency, that is, without causing large head losses.  The head used in the calculations is 400 feet.

Not surprisingly, the analysis reveals that the capital cost per unit of power production decreases with increasing pipe diameter.  Since the pricing of the turbine assembly varies in a similar fashion, the lesson is that the greatest investment efficiency may be achieved in the largest installation that can be undertaken within the budget available for the project.  This is particularly the case in that larger pipe sizes can accommodate higher flow velocities and still remain in the efficient realm of the flow regime, meaning that the power production estimates for the 12” and 15” pipelines below are thought to be conservative estimates.  On the other hand, the

Table 1.  Capital cost / benefit comparison of different pipe sizes based exclusively on materials cost, assuming installation costs to be comparable.

labor for installation of a 15” diameter pipeline is somewhat of a greater challenge and more costly than the installation of an 8” diameter pipeline, so there is likely to be somewhat of a cancelling effect in the divergence of reality from the assumptions used in the ballpark comparison.

ECONOMIC ANALYSIS

In order to undertake a preliminary economic feasibility assessment, a system design comprised by 15” pipe was employed.  This system presupposes 400 feet of net head and 2800 gpm of flow across a one-kilometer length of river, which corresponds to a hydroelectric potential of 131 kilowatts.  A considerable number of variables that remain undefined makes it difficult to pinpoint costs, though Table 2 provides a ballpark first stab estimate.  Some of these factors are listed below:

1)      Power Transmission Requirements.  It will be necessary to tie into the grid at a point at which the power grid electrical cable is adequately sized to transmit the power being sold back to the grid without unacceptable line losses.  It is likely that some capital costs will be required to provide transmission from the generation point to the point of grid connection.  It is possible that the power company will be willing to assume all or part of this investment, and if not, specifications for the grid extension will be needed from ICE in order to approximate capital costs of the required grid extension.  Transmission capital costs were not factored into the present analysis.

2)      Turbine Housing.  A structure will be required to contain the hydroelectric generator and provide a discharge of the penstock water back to the Río Negro after passage through the turbine.  Dimensions and specifications of this structure will naturally vary as a function of the size of the plant ultimately deployed and also according to supplemental functions that may be satisfied by the building.  An arbitrary cost of $25,000 is assumed to be an adequate budget for such a structure.

3)      Intake Structure.   Hydroelectric intake structures can vary dramatically in scope, scale, and cost.  In the case of the Río Negro, the dry season flow rate will be a strong factor in determining the optimal intake point(s).  A small reservoir would represent a significant capital investment and dramatic environmental disturbance, whereas an infiltration gallery or comparable in-stream or adjacent intake capture system (like that deployed at the base of the property for municipal water intake) would be considerably less capital intensive.  For the purposes of this preliminary analysis, a capital cost of $15,000 is projected for the intake structure.

4)      Installation.  The cost of pipeline and turbine installation is another variable that is difficult to project at this stage, requiring considerable additional engineering analysis and consultation with subcontractors before arriving at a reasonable estimate.   The figure of $75,000 used for preliminary feasibility evaluation should be assumed to be a rough stab in the dark as a starting point.

5)      Inflation.  The price of steel and plastics are in a period of dramatic price inflation due to volatility of the global oil market and increasing demand in developing nations for steel, particularly China.  Besides volatility in materials cost, the inflation in construction costs is a non-trivial consideration.  Taken together 20% annual inflation in project costs is not an unreasonable expectation.

6)      Transport and Import Tariffs.  It is possible that import duties may be waived for the hydroelectric plant, or alternately it may be possible to negotiate reduced tariffs since the installation is to have a direct benefit to the nation of Costa Rica.  However, if import duties are not waived, then 30% tariffs will represent a significant additional cost.  Similarly, shipping of parts varies dramatically.  A figure of $10,000 is used to approximate shipping and trucking costs, but that is purely ball park.  No tariff surcharges were used in the cost estimation.

Table 2.  Ballpark estimate of capital costs of 130 kilowatt hydroelectric power plant.

Current domestic rates for electrical power are $0.10 per kilowatt-hour.  Commercial rates vary as a function of usage, but I paid $0.16 per kilowatt-hour on average in 2007 for commercial service to an Internet Café.  If we assume power sales back to the grid at 75% of the value of residential kilowatt-hour pricing, a 130-kilowatt facility will have a hypothetical revenue stream of $234 per day, or $85,410 per year.  Abstracting for the moment annual maintenance costs, the project has a payback period of 5.56 years. 

Assuming a 20-year amortization and discounting such factors as the increase in the value of electrical power and annual maintenance, the total revenues point to 359% profit margin over twenty years, or 17.9% annual return on investment.

CAVEATS

The economic analysis summarized above is preliminary, with many ballpark estimates and a number of considerations that have not been factored into the analysis at all.  A number of caveats are therefore worth mentioning.

1)     Permitting/Legal.  The fees associated with obtaining a water extraction permit and of negotiating a contract with the government of Costa Rica to buy back the power produced are likely non-trivial and even if they do not represent a dramatic capital sink, they will delay any actual project groundbreaking so that higher costs than initially projected are almost a certainty due to inflation in materials and the cost of construction in general.

2)     Inflation in Energy Prices.  The other side of the coin is that the end product, electrical power, is one that is also expected to continually rise in value, so that any contractual agreement must be predicated on a scale of remuneration that slides according to the price of electricity.

3)     Internal Consumption.  While the economic model was predicated on a purely hypothetical value of the power produced of 75% of its residential sale price to the public, the system can also be used to provide facility electrical power demands so that savings in power consumption can be factored into a global facility cost/benefits analysis as revenues adduced from savings in power costs.

4)     Actual Project Life Expectancy.  In reality, the facility will have a lifespan that considerably exceeds the 20-year project life period used in economic calculations.  Though replacement costs of turbine runners and other parts are expected to be non-trivial expenses on intervals of several years, the facility should be able to expect a reasonable life span overall of several decades.

5)     Intangibles.  Value cannot be universally equated with a monetary sum, so the added value of doing the right thing for the planet is one that is difficult to factor into an economic analysis but which is nevertheless a significant added value.

CONCLUSIONS AND RECOMMENDATIONS

The Río Negro offers an ideal setting for the installation of a high-head, low-flow mini-hydroelectric grid-tie power generation plant.  Because Finca Adiabet bounds the river for two miles across a vertical relief of as much as 2000 feet, there are a variety of physical configurations in which hydroelectric power could be harnessed.  AC-Direct hydroelectric power that is produced must be used, however, or burnt off as heat, so all the hydro potential in the world is of little value without being able to tie into a local or regional power grid that can absorb power above that required for internal consumption.  Consequently, the first step in determining if a mini-hydroelectric facility is feasible at all is the negotiation of favorable terms for the long-term sale of the power to the national power company.  The terms of this agreement should strive for the following governmental accommodations:

1)     A purchase price that is tied to the retail price of grid power so that as power prices increase, the Adiabet Hydro facility will benefit from commensurately increased revenues.

2)     An agreement to waive import tariffs on the steel pipeline (if steel is selected as the pipeline material) and the hydroelectric power plant itself, which of course must be imported.

3)     Governmental assurance that no sunset clause can be invoked that might subsequently impact the economic model justifying the original investment.

4)     Possible mediation to expedite water concession permitting on terms favorable to the resource exploitation.

Simultaneous to the application of an agreement with the government to buy the power, the scope of the project must be defined in conjunction with MINAE and Department of Waters personnel to determine the flow rate that will be permitted for hydroelectric diversion.  While it is easily argued that the water is not removed from the environment since it is returned, it is very likely that regulators may see it differently and argue for a ceiling on diversion flow rates to ensure that healthy dry season flows are sustained in the river.  The flow rates that are ultimately negotiated are almost certainly the second most important variable to ensuring the project viability following the governmental commitment to buy the power in the first place.

AC-Direct systems, unlike micro-hydroelectric applications, do not have a great tolerance for variation in flow rates.  Therefore while it is reasonable to anticipate somewhat greater production capacity during the rainy season, the overall system must be designed based on the base flow that can be expected throughout the whole year.  In other words, the final design is likely to be most rationally based on the flow available at the driest time of the year.  That is a variable that must be determined next March or April and must be done in collaboration with permitting officials that are ultimately responsible for approval of the water concession that will permit such diversion.  Having said that, wet season flow rates are dramatically higher, and at some finite project scope, it may be economic to have a large-scale hydroelectric plant that operates only during the rainy season that capitalizes on dramatically increased flow rates.  While this is a theoretical possibility and one that should be explored during further project viability studies, it is unlikely that a seasonal application will be the most economically favored alternative.

Correspondingly, while a proposal is submitted to ICE and the Controlaria de la Republica to sell power to the grid, a water concession application should also be initiated.  The first step of that process is the completion of an environmental impact statement for the governmental body SETENA to ascertain whether a more comprehensive environmental impact study must be conducted or whether that possible ancillary requirement may be waived.  It is only following SETENA’S approval of the environmental impact assessment (whether as a statement or a study) that a water concession can be solicited from MINAE and the Department of Waters.  Therefore, these steps should be taken as soon as possible to ensure that the process is sufficiently advanced so that inspectors can be dispatched during the 2009 dry season to make final water diversion decisions.  By missing the 2009 dry season window of January-April will put the project off for another year.

The whole development process is iterative and inter-related.  It is not possible to finalize a deal with the government to buy back power without knowing how much power that will be.  And it is not possible to know how much power can be made without knowing how much water will be permitted and at what location.  To confound matters, capital planning cannot proceed without understanding the scope of the project either.  As a consequence three inter-related jobs must be undertaken at the same time and coordinated to update one another with relevant feedback.  Those three project development thrusts are the following:

1)     Water concession and extraction permitting.

2)     Governmental contract to buy back power

3)     Engineering, design, work plan, delivery timelines, and budgeting.

Appendix 1:  Field Notes