Resource Evaluation:  Finca Cerro Osa

Osa Peninsula, Costa Rica

For:  Cerro Osa Ownership

By:  Paul Collar, Osa Water Works, SA

March 18, 2006

View from existing house over future cabina and home sites

 INTRODUCTION AND OBJECTIVES

The spectacular Cerro Osa property that is the subject of this report comprises terrain descending from a ridgeline to the flanking valleys and ridges and has a wide range of elevations and multiple perennial streams, a river, and myriad springs across the property breadth.  This property is inscribed in a Costa Rican private land-zoning governmental conservation program that makes most of the property practically equivalent to national parkland.  The allocated sustainable development areas in such programs are typically restricted to around 10% of the total property acreage, and in the primary ten-acre highland clearing permitted for development at Cerro Osa, the ownership envisions a complex comprised by several bungalows and one large guest house and three additional homes to collectively comprise a remote ecolodge resort and residential complex.  Additional residential building sites are being considered outside of the central 10-acre development that is the focus of this report. 

The objective of this study was to determine the most effective use of property resources to supply water for potable, irrigation, and pool water needs as well as to determine the viability of hydroelectric power generation to supplement existing solar and fossil-fuel power generation capacity.

ANTICIPATED FACILITY DEMANDS

Water demands for the facility include potable, irrigation, pool-water, and hydroelectric power components.  In this section, the first three water requirements are discussed and hydroelectric potential is reserved for discussion in a subsequent section.

Potable Water Demand.  Given a projection of 8 cabinas with a peak capacity of 4 persons per cabina, a guest house with an expected peak occupancy of 16, a main house with an occupancy of an additional four and five on-site staff members, that comes to a total peak occupancy of around 60.  The American residential daily consumption criterion used for water supply design is 100 gallons per capita.  For a Costa Rican ecolodge the actual potable water demand per full-time guest is expected to be considerably lower than this, likely less than half this amount.  However, using an estimated maximum consumption of 75 gallons per day per capita, the facility would have a peak water demand of 4500 gallons per day, a flow rate that corresponds to slightly more than 3 gpm.

Pool Water Demand.  Pool water demand consists of evaporation and splash losses that must be made on a daily basis.  Evaporation rate is a function of surface area and therefore pool geometry.  However, presuming exaggerated losses of 1% per day and a pool size of 50,000 gallons, pool water demand would represent an additional 0.34 gpm in water supply requirements.

Irrigation Demand.  Irrigation water demands vary as a function of crop type (cover, row, greenhouse, vs. orchard), irrigation style (sprinkler, spray, drip), watering schedule (evaporation losses are greater during the day), plant species, soil characteristics, and other factors.  A rule-of-thumb irrigation design criterion for the equatorial tropics is the application of 5 mm of water during the dry season.  If we assume that the entire ten-acre complex will require irrigation, this implies a daily irrigation water demand of 54,000 gallons applied over the 12 hours of darkness, which corresponds to an operational irrigation application rate of 75 gpm.  In order to sustain this flow across twelve hours based on a 40 gpm feed rate requires a storage tank with a capacity of 26,000 gallons.

Facility Power Demand

The Cerro Osa ownership has compiled an estimation of anticipated facility power requirements, which are shown in Table 1.

Table 1.  Cerro-Osa supplied energy demand calculations.

The analysis of appliance draws, usage patterns, and corresponding power demands suggest, therefore, a 16 kw daily power demand. 

PROPERTY RESOURCES AND HYDRO OPTIONS

The area in the vicinity of the 10-acre development site has a spring ideal for potable water supply and two watersheds with perennial streamflow and significant relief in reasonably close proximity, one on either side of the development site.  Of these two drainages, the named Quebrada Coyunda is the most appropriate for hydroelectric and irrigation water supply development owing to its significantly higher flow.  Figure 1 shows a site map with portions of the property relevant to this resource evaluaton, and Figure 2 is an energy diagram showing elevation relative to the existing house site. 

Figure 1.  Map view of portion of property map showing the development zone and nearby water sources.

Figure 2.  Energy diagram showing relative elevations of locations relevant to this investigation.  The datum was established as the elevation of the existing house.

The domestic potable water water for the main house is presently achieved by the collection of surface water in a makeshift water intake structure at the confluence of one ephemeral stream and a second perennial stream fed by about 1 gpm deriving from a spring about 100 feet away.  Because the existing water supply is not protected and is exposed to natural contaminants (see photos below), the uncontaminated water emerging from the spring accumulates contaminants in the collection structure; water from the kitchen tap had a fecal coliform concentration of 17 colonies / 100 mL. 

The unnamed stream that is formed from the spring increases in flow downhill as it drains the property adjacent to the development area, and in the region of the cuidador’s house had a sustained dry-season flow of around 25 gallons per minute.  However, this surface water source is located at an elevation lower than the building sites and its sustained dry season flow is a mere fraction of the flow of Quebrada Coyunda, one drainage over.  The unnamed stream was therefore discarted as a likely source for water supply or hydroelectric power generation.

At the Coyunda confluence of the Left and Right Forks, dry season flow was estimated at considerably higher than 1000 gpm.  Upstream in the Left Fork, a waterfall allowed a measurement of stream flow using a watch and five-gallon bucket, yielding a measured 400 gpm and estimated 50% losses for a flow approximation of 600 gpm.  At the Coyunda confluence the left fork appears to comprise 60% of the total flow and the right fork 40%

The Left Fork of the Coyunda had an intake location 230 feet above the confluence where the extraction of 60 gpm was possible.  Above this, the stream declined in flow rate rapidly.  If the Right Fork is hydrologically comparable to the Left, the two tributaries together may be able to contribute 100 gpm to hydroelectric diversion at this approximate elevation.  To reduce power transmission distances to a minimum, it will be optimal to generate 2-300 meters downstream of the Coyunda confluence, where the stream is closest to the development site.  This distance is expected to contribute an additional 50 feet of head, though this requires confirmation in the field.    A dual-pipeline high-head low-flow (HHLF) hydro system (Figure 3) would have a daily power generation output, based on 100 gpm and 280 feet of head of 2.52 kw and a daily 60 kw charging source, more than three times the energy demands anticipated by the ownership in Table 1.  A considerably simpler HHLF configuration alternative is to abstract the hypothetical Right Fork wing of the pipeline from the design (Figure 4).  Based on confirmed field observations of 50 gpm and 280 feet of head, the hydro potential of the single pipeline HHLF alternative is 1.26 kw, or 30 kw per day.

Figure 5 shows a low-head, high-flow (LHHF) configuration with an intake at the confluence of the two forks.  Power generation comparable to the HHLF dual pipeline alternative is theoretically achieved across a 50 foot elevaton drop through a diversion of 560 gpm of water (6” pipe).  At the time of the survey, the flow rate at the confluence was estimated to be in excess of 1000 gpm.

It is also possible to achieve a moderate-head, moderate-flow alternative in the right fork tributary (Figure 6), where diversion of 250 gpm (slightly more than 50% of the flow) for a total head of 100 feed for power generation 50 feet below the confluence would provide for 2.25 kw (54 kw/day) and would require a 3” diameter pipeline.

In addition to the raw untapped energetic resources of the Coyunda watershed, Cerro Osa has a solar power system in place and under expansion.  The fact that battery capacity, AC inversion, and associated electrical systems are already in place for solar power removes one of the primary capital expenses of micro-hydro applications.  Beyond existing and planned solar capacity and proven hydroelectric potential, the property enjoys a physiography that may be particularly conducive to wind-generation, though feasibility studies for this will necessarily involve regular collection of wind-speed measurements at various heights above the land surface for optimal design.

ENGINEERING ANALYSIS

Irrigation water supply

A hydraulic ram pump depends on the kinetic energy of falling water and two moving parts to pump water; it is proven, time-tested technology requiring minimal maintenance and energy-free continuous operation.  It’s rule-of-thumb design variables provide for lift ten times higher than the elevation difference between the drivepipe intake and the pump, up to a practical total of 500 feet.  It has a yield of approximately 1/10^th of its input.  In order to achieve a continuous flow of around 40 gpm to satisfy the Cerro Osa irrigation water demand, the most appropriate location for deployment of a ram pump is in the Left Fork of the Coyunda from the base of the waterfall shown in Figure 7.  This location is about 300 feet below the optimal elevation of an irrigation storage tank.  At a 6:1 drop:lift ratio (presuming 50 feet of driveline head), water delivery is actually closer to a 1:8 ratio, which means that to deliver 40 gpm requires a feed water flow rate of 320 gpm, which presupposes a 4” diameter driveline.  Presuming a 25% average grade in this steep part of the watershed, the driveline length will be about 200 feet.  Owing to the heavy water hammer developed, galvanized steel is the material of choice for driveline construction.  A conceptual ram pump/storage tank design is shown in Figure 8.

While there is a significant pricing latitude among different manufacturers, most off-the-shelf ram pumps are for much smaller applications.  A cursory Internet review showed the largest priced pump identified during a brief surf to be a 3” pump offered for $2100, so it is reasonable to anticipate that a 4” pump, the minimum required for this application is likely to run $3500 for a knock-off rather than a Vulcan, the original ram pump manufacturer and industrial standard-bearer of this technology.  In order to irrigate 5 mm across 10 acres during twelve hours daily at a 40 gpm ram pump delivery rate additionally requires storage capacity of 26,000 gallons. 

Given the high anticipated costs of the system, the following analysis is presented to determine if an electric pump may be deployed for the service.  This analysis first requires an assessment of the energetics.  Using a Total Dynamic Head (the sum of static head, pressure head, suction head, and friction losses) of 350 feet and the desired pumping rate of 75 gpm (the actual irrigation delivery rate required) the Water Horsepower is found to be 6.6 HP and division by pump efficiency values puts a Brake Horse Power over 7 HP.  This is approximately equivalent to a 5 kw power demand for operation twelve hours daily for a total hypothetical pumping demand of 60 kw per day.  This is the upper end of the stream’s 24-hour hydroelectric potential, and AC pumping would therefore absolutely require fossil-fuel AC generation to satisfy this power demand. 

Pool water supply.  Pool makeup water will be most efficiently derived from the irrigation water supply.  However, expanded storage capacity in the potable water system will make it so that makeup water can be derived from drinking water except, arguably, under peak occupancy. 

Potable Water Supply.  The existing spring was slightly exposed in order to capture a water sample.  In five minutes of playing in mud, I was able to produce ½ gpm from the largest vein of water evident.  It is my professional estimation that full development of this source will increase the captureable flow to 2 gpm.  This adds up to 2880 gallons per day.  Recalling the peak occupancy figure of 60, this amounts to 48 gallons per day per person.  In practical terms, it is unlikely for a person to use this much water for showering, laundry, ingestion, cooking, and cleaning.  Also, water-conservation toilets, faucetts, and sinks can provide additional economy.  It is very likely that the existing water supply, provided adequate water storage is included in the system design, will suffice for the facility needs, even during peak occupancy.  Pool makeup water is expected to be available except during peak occupancy. 

Nevertheless, best design practices would presuppose the irrigation tank at a higher elevation than the potable tank in order to be able to compensate potable shortfalls during peak occupancy.  An inline particulate removal and disinfection capacity would ensure that added water is entirely potable.

RECOMMENDATIONS.

Potable water supply.

The existing potable water supply should be replaced with a spring box like the one depicted conceptually in Figure 9.  While the current pipeline is only ¾”, it will make most sense to eventually deploy a 2” trunk pipeline to reduce friction losses and provide operational capacity for peak usage demands.  The flexibility of the system is ultimately dependent on storage capacity as well, and this can be supplemented at will by simply plumbing additional plastic tanks into the system, or constructing a larger concrete buried storage tank 100 feet above the development site of a volume best determined following spring box completion and measurement of the real-world water supply flow rate.  Also, storage can be factored into the spring box itself.  For instance the 2 m x 2 m x 2 m spring box proposed provides 2200 gallons of storage whereas a 1 m x 1 m x 1 m spring box provides only 264 gallons of storage.

Figure 9:  Spring Box conceptual design.  Please note that the system capacity presumes daily water demands of 100 gallons per resident, a typical value used for planning residential American water supply requirements.  In practice Costa Rican residential water demand is significantly lower, and per capita demand in a commercial installation, like a lodge is lower than for domestic consumers.  The system is capable with adequate storage of providing 40 gallons per day to each of 70 people under peak occupancy.

Osa Water Works standard design and installation practice is to cloak intake structures, tanks, and pipelines wherever possible.  In this case, I propose a spring box mostly buried into the slope where the spring is located, with an internally located four inch clean-out valve, an access hatch, an overflow routed to emerge as a faux spring beneath the base of the box in the stream channel itself, and the spring box face stuccoed with rock.  A budget for spring potable water supply installation is given in Table 2.

TABLE 2.  Installation estimate for spring box (2 meters cubed) and connection to existing water storage and distribution system.

Hydroelectric

The property presents four alternative hydroelectric configurations in the Coyunda watershed.  These are summarized below.

1)     Dual Fork High Head Low Flow Max:  2.5 kw (Figure 3)

2)     Left Fork High Head Low Flow:  2.2 kw (Figure 4)

3)     Left Fork Moderate Head Moderate Flow:  1.2 kw (Figure 6)

4)     Quebrada Coyunda Low Head High Flow:  up to 2.7 kw (Figure 5)

The fourth alternative is within a hydraulic regime poorly suited to micro-hydro applications.  Consider that at 50 feet of head a four-nozzle Harris Pelton Wheel can only process a total of 100 gpm, which means it would require six pelton wheels in parallel to achieve a generation rate comparable to the Option 1 dual-pipeline HHLF alternative.  Under such a phalanx of turbines, each Pelton would be producing only 450 watts, or less than one third of its 24-Volt potential.  Other turbine types more efficient for low head, high flow applications are usually associated with AC-Direct power generation rather than micro-hydro applications.  While a 2.7 kw AC-Direct alternative is intriguing, the extensive existing commitment to solar power generation and associated electronics and battery bank strongly favors a micro-hydro application that can be integrated with solar, wind, and fossil fuel power generation at a central energy center.  For this reason, the LHHF option was dropped from additional consideration.

While a variety of micro-hydro turbine/generator assemblies are available on the market, each with its own specs and design implications, Harris Hydroelectric remains the standard-bearer of the industry, and consideration of the remaining three hydro configurations uses the operating ranges of the Harris pelton wheel.  Voltage transformation to 480 V at the generation site and back to 120 V at the Power Center is presupposed for all the alternatives in order to transmit power at a modest cable gage without unacceptably high line losses.

Both HHLF alternatives have hydroelectric potentials in excess of the 24-Volt capacity of a single Pelton Wheel.  The moderate head moderate flow (MHMF) configuration, on the other hand, at 1.2 kw, is just inside the capacity of a single Pelton wheel.  While the HHLF alternatives offer considerably more power, they each require a second turbine to capitalize that poential.  This in turn invokes greater costs in the enclosure size in addition to a manifold.  However, the two HHLF alternatives additionally allow for maximal rainy season hydroelectric utilization, when hydro potential may be reasonably expected to be consistently 50% higher than the dry season expectations, possibly more.  This may be an important consideration, given that hydro has maximal capacity at a time when solar is at its minimum through climactic suppression of insolation.  The MHMF option, at 250 gpm intake, is already close to its peak practical flow capacity in three inch pipe.  To introduce rainy season expandability would require that this pipeline be deployed in 4” diameter pipe.   On the other hand, the two inch pipe in the high head alternatives is under-utilized at the dry season flow rate of 50 gpm and dramatic gains will be seen in the rainy season that allow for expansion with additional turbine(s) and not require pipeline upsizing.

In order to determine firm costs, actual pipeline distances must be measured in a follow-up survey.  While approximate distances of the portions of the watersheds visited are summarized in armchair estimation for the three hydroelectric alternatives considered most viable (Tables 3 through 5), it is emphasized that the budgets presented include estimated distances and should not be considered final firm costs.  The final line of each graph contains a cost/benefits parameter equal to the capital cost divided by the daily hydro potental in kilowatts.  Using strictly cost/benefits criteria for economic decision-making, the MHMF alternative is the most attractive.  However, it is also the alternative least tolerant to expansion of generation capacity during rainy season months, whereas both the dual HHLF and Right Fork HHLF alternatives offer robust rainy season capacity expansion because of the fact that 2” pipeline can accommodate flows significantly greater than the 50 gpm each is expected to carry during the dry season.

Table 3.  Estimated Installation Budget for dual pipeline HHLF configuration (Figure 3).

Table 4.  Estimated Installation Budget for single pipeline HHLF configuration (Figure 4) with the right fork eliminated and the using two inch pipeline the entire length.

Table 5.  Estimated Installation Budget for Moderate-Head Moderate-Flow configuration (Figure 6)

Irrigation

Estimated installation costs for the ram pump / storage assemblage shown in Figure 8 is given in Table 6

Before any decisions are made regarding an irrigation system, and since there is no need for such a system this year, the ownership may wish to studiously assess actual irrigation needs.  The 5 mm per day across 10 acres is a very broad rule-of-thumb Table 6.  Estimated costs of ram pump / irrigation storage tank system.  Please note that the time estimate is for installation of pipelines, intake and pump.  Manpower is already factored into the rule-of-thumb pricing for tank construction ($1 per gallon of storage).  Tank construction will take an additional 3-4 weeks.

irrigation design criterion, and actual water requirements will vary as a function of irrigation style, the plant being watered, the time of day of the watering and also on operational decisions.  There are thousands upon thousands of fruit trees on this peninsula that have never seen a drop of irrigation water and are doing just fine.  Similarly, the ownership may elect to tolerate water stresses on the grass during the hot summer months and husband irrigation water for gardening, greenhouse, saplings, and similarly vulnerable plants.  On the other hand, the irrigated areas may be constrained to a much smaller area and require less total volumes with some downward adjustment of the system capacity and associated capital costs.

If the ownership should ultimately elect to provide for the scale of irrigation discussed in this report, the most practical solution may be to use a small tank (5000 gallons) and a 10 HP AC submersible pump and intake housing bored into the river bed adjacent to the hydro generation point.  This alternative would carry one fourth of the capital costs of a ram pump system and have a much smaller environmental footprint (this system can be completely invisible to an unpracticed eye).  The downside is dependence upon a non-renewable fossil-fuel resource to power the pump. 

Conclusions

Additional field work will be required to determine distances and for siting the optimal generation point on Quebrada Coyunda.  Similarly, a straight-line trajectory and distance measurement from the generation point to the development site is required for a final scrutiny of transmission efficiency and final costing.  The projected costs presented in this report should be considered estimates, as they are based on approximated pipeline lengths.  Nevertheless, Cerro Osa possesses robust hydroelectric alternatives either for a stand-alone power system or for integration into a hybrid power system with solar-, fossil-fuel, and optionally wind-charging sources.  The existing spring, with adequate storage capacity, is expected to satisfy facility potable demands even at peak occupancy with pure water requiring no subsequent treatment.  Finally, the nearby Coyunda tributaries contain sufficient dry-season flow to provide for dry-season irrigation needs, though the capital costs of a ram-based system is a bit daunting at the scale of irrigation demands assumed for this report.

Headed home after another long hard day at the office.