Osa Water Works Resource Evaluations: May 6, 2008

View from Amelia: Osa Peninsula , Sierpe and Terraba mangrove swamp, and Pacific Ocean

Amelia Gardens is a proposed residential community located near Cortés and Palmar Norte in the Pacific coastal mountain range highlands overlooking the great Sierpe mangrove swamp. The extensive development plans for the 700-acre property include residential, commercial, and educational aspects, all predicated on best management practices for environmental sustainability. Osa Water Works (OWW) was contracted to provide a resource evaluation and highlight water, energy, and waste management alternatives during the dry season to support the overall project scope.

Technical work that has been undertaken to date that was shared with OWW includes a large-scale topographic map (Figure 1-A), an architectural land-development master plan (Figure 1-B), and a topographic rendering of the architectural plans developed to date (Figure 1-C). A description of the scope of the project has been presented in a series of emails and in the master plan. Key parts of the development plan are bulleted below:

Project planners intend for Amelia Gardens to be an example of environmentally sustainable development to include alternative energy capitalization, rainfall capture, water-resource optimization, appropriate-technology wastewater management, and both perma- and aqua- culture demonstration projects.

Costa Rica is blessed with an abundance of rainfall. But rather than being nicely distributed this rainfall occurs mostly during eight months of the year. The dry season (December-April) is also the prime tourism season. This means that not just for Amelia Gardens but for all residential, commercial, and related projects dependent upon the patronage of expatriate clientele, that the period of peak usage coincides with the period with the least amount of available water. Since Costa Rica depends on large-scale hydroelectric for 86% of its domestic power needs ⁽¹⁾, this means that the dry season months represent the baseline prime design criterion for both water supply and overall electrical power sufficiency, not just for onsite mini- and micro-hydroelectric viability, but indeed, even for national grid hookup. In 2007 the arrival of the rainy season two weeks late caused nationwide rolling blackouts as hydroelectric reservoirs lost capacity because of low water levels ⁽²⁾.

Figure 2. Rainfall distribution for San Jose; note demarcation between dry (Dec-Apr) and rainy (May-Nov) season ⁽³⁾.

If demands of the dry season can be met at Amelia Gardens as in Costa Rica in general, then the remaining eight months of the year can be reasonably assumed will fall into place.

In the case of Amelia Gardens the dichotomy between dry and wet season stream flow rates is likely to have a dramatic impact on overall project viability as currently envisaged. Much has been made about rainfall capture and about rainy season hydroelectric as ways to provide measures of environmental sustainability. However, the cruel irony is that periods with the greatest amount of available water coincide with periods of lowest expected occupancy and energy demands. Until the Republic of Costa Rica embraces power purchase from small-time power generators, it will not be practical for facilities like Amelia Gardens to consider a dramatic capital investment in onsite AC-power generation (whether wet season hydroelectric, by wind, or by municipal-scale solar power farming). This is the single most important national policy evolution that is required by the government to be able to minimize the anticipated power shortfalls that are anticipated by the expected 6% annual growth in power demand, coupled with the changes in rain distribution expected to accompany climate change. Already there are rains in the summer and dry spells in the winter, something inconceivable here less than twenty years ago.

In an analogous manner, the concept of rainfall capture for potable water supply is certainly applicable for residential domestic demand, with the limiting design variable being the storage capacity required to get through four months of dry season ⁽⁴⁾. Perhaps changing climate patterns may allow for designers to be able to include mid-summer rains in a rational facility design, but conservative planning would necessarily be predicated on the traditional expectation of four months without any rain whatsoever. And under these conditions, rainfall capture is viable for domestic potable water supply mostly nationwide but is tenable for neither irrigation, commercial, nor industrial water supply anywhere.

This study was intended to determine the extent to which property resources can be harmoniously developed to support project water, energy, and waste management requirements of the Amelia Gardens architectural master plan.

United States design criteria for municipal water systems allow a 100 gallons per person per day for each household occupant. In Costa Rica, the number used for the same purpose is closer to 75 gpd per person. Using Costa Rica’s number and presuming three full-time residents per casita household and 2 residents per cabin household, this amount to 400 persons at full occupancy. Adding in another 100 persons for peak occupancy at the condominium hotel and to account for facility staff amounts to a projected peak facility occupancy of 500 persons and a daily potable water demand of 37,500 gallons, which is equivalent to a continuous flow rate of 26 gpm.

Projected casita power demand is estimated in Table 1, which reveals that with all the appliances on at the same time, a house can be expected to have a peak power demand of up to 8 kw. Since there

Table 1. Power estimation of individual casita presuming conventional appliances.

are 100 projected casitas, this amounts to 800 kw if everything is on at once. In practical terms this is unlikely, so a more rational peak projection at full occupancy would likely be closer to 500 kw. If we assume that the 50 cabins have fewer amenities and have a peak power of 4 kw, that adds 200 kw, or practically speaking 150 kw at peak. In addition to the 650 kw in peak residential power demand, it is reasonable to assume that infrastructural power demand may add another 25%. Since commercial parts of the facility (excluding completely for the moment the condominium hotel) are expected to have peak power requirements during the day when the residences are not at peak, then it is likely that decreased residential demand may provide for commercial power demands during the day so that these can be momentarily accounted for. Adding the 25% for infrastructural power requirements suggests that a facility-wide peak capacity of 815 kw should be anticipated as a ballpark baseline peak power demand for residential and facility operations. With peak capacity from the condominium hotel (with its presumptive air conditioned units), the peak power demand is likely to approach 1.5 megawatts.

The author of this report was guided by Freddy Jimenez to all portions of the project with surface water resources. The field investigation was undertaken on the 23^rd and 26^th of April, 2008, at the very driest time of the year. Water flow rates were visually estimated, and elevations were measured using a barometric altimeter. Locations were determined using a handheld global positioning system. Figure 3 shows the GPS track of the hikes taken through the property overlain on the large-scale topographic map; Figure 4 shows the track overlain on the rendered topo, and Figure 5 shows the same information overlain on the architectural master plan. Scale adjustments were based on the shape of the curves of the road and the location of the lowest waterfall and should not be taken as exactly precise overlays, though reasonable matches for the purposes of this report. Raw field data are given in Appendix 1.

Figure 3. GPS field survey track superimposed on large scale topographic map.

Figure 4. GPS field survey track superimposed on rendered topographic map showing architectural features.

Assuming a peak occupancy of 500 persons, the conventional Costa Rican per capita water consumption design criterion of 75 gpd presumes a peak daily water demand of 37,500 gallons. This is equal to 26 gallons per minute (gpm) of continuous flow. This means that if adequate storage capacity is factored into the system design, then a flow of as little as 26 gpm is capable of providing full potable and light commercial water demands during full peak occupancy of 500.

The housing and commercial facilities at Amelia Gardens are planned at a variety of elevations. To provide for gravity water supply for all from a common source requires that the highest installation represent that datum upon which the water intake location is based. The highest point of the project is the Condo Hotel site. Optimal water pressure is 40 psi, so for optimal gravity feed to the condo site, an intake and storage is needed at least 120 feet higher.

There is only a small reach of the river that can spare this amount of water at this elevation without exceeding 20% of the total dry season base flow; the best intake spot is shown in Figure 6-A. For a better visualization of the estimated 75 gpm flow rate, the photo of the waterfall (Figure 6-B) was taken about 200 meters downstream.

Figure 6-A. Proposed intake point

Figure 6-B. Waterfall 200 meters downstream of proposed intake point.

While it is possible also to imagine a series of smaller water systems to provide mini-systems along the property, the most practical overall water-supply solution is based upon a single surface water intake high in the watershed. A large tank and large diameter distribution main are essential for this design approach, shown in Figure 7.

Figure 7. Potable water system design comprised of water intake by infiltration gallery, 100,000 gallon storage and pressure tank, 12” water mains, and 6”, 4”, and 2” secondary mains. Individual housing stubs are not shown but should be completed in 1” pipe with interior home plumbing in ½” D.

Rainfall capture is arguably the most environmentally sustainable form of domestic water supply. It consists simply of capturing rain in a large storage tank and withdrawing from the tank for potable uses. The design process requires analysis of rainfall distribution and intensity in order to provide adequate capture surface and storage design variables. The intensity of rainfall during the rainy season is such that any modest-sized house provides enough roof area to ensure adequate capture, so in practice the controlling design variable devolves to tank size. To arrive at this, I have assumed an occupancy of four permanent residents and a daily water demand of 75 gallons per person. Since there are four months without rain, this means that for a family of four the amount of water that must be stored is 36,000 gallons. The only practical place to situate a tank of this size is as the foundation of the house itself. Supplemental requirements are a pump and pressure tank ($2000 installed) and a home filtration system that includes particulate, granular activated carbon, and ultraviolet disinfection ($2000 installed). As a rule of thumb, tank costs run $1 per gallon of storage, so the costs of outfitting each house with its own stand-alone rainfall capture potable water system is around $40,000. On top of this, the power draw from the one horsepower centrifugal pumps required for pressurization of rainfall-capture homes add 1-2.5 kw per day of power demand to each dwelling as the energetic cost of pressurizing domestic water.

The hydroelectric potential of the perennial stream reach within the property, including 600 feet and more of vertical relief, is enough to sustain 3-4 modest homes with micro-hydroelectric power generation. In AC-Direct, however, the dry season flow rate is expected to be able to sustain the full time power needs of only a single casita home.

Hydroelectric potential is defined by the empirical equation given in Equation (1)

Where P is Power in watts, H is head in feet, Q is flow in gallons per minute, 0.18 is a units conversion coefficient, and e is efficiency (0.5 for micro-hydro systems, 0.65 for mini-hydro systems).

An estimated flow of 75 gpm was observed at Location 590 (Figure 7). A waterfall at this location provided a reasonable location upstream for a water intake. Whereas 75 gpm was estimated within +/- 20% to be the total flow of the stream and therefore not useable in its entirety without adverse environmental and aesthetic impact, let us presently assume that the entire flow rate is captured and used to turn a hydroelectric turbine at Location 599 at the base of the large waterfall. This corresponds to a vertical relief of 607 feet. For a mini-hydro AC Direct system, this corresponds to a dry season hydroelectric potential of 5.3 kilowatts. This is nearly enough power to satisfactorily power a single casita based on the power demand projections summarized in Table 1, except that the system would permit only about 2/3 of facility appliances to operate simultaneously.

In other words, the hydroelectric potential is minuscule in comparison to the anticipated demands of the overall project. To complicate matters, it is environmentally unsustainable to channel the entire stream flow into a pipe to generate power. Because seeps occur throughout the river channel, it is arguable that 50% withdrawal in the upper watershed is indeed environmentally sustainable since it is continually replaced moving downstream by the abundant seeps that constitute the stream’s dry season base flow. However, Costa Rican regulatory bodies are unlikely to permit more than 20% of the flow for diversion for either potable or hydroelectric power generation purposes.

Ignoring for a moment that potable water supply already potentially calls for as much as 20% of the upland stream flow rate, a 20 gpm flow rate over a head of 607 feet provides a mere 1.1 kw of continuous power potential. Since micro-hydro presupposes power storage in batteries, the 1.1 kw potential in practice amounts to 26 kw per day. Therefore, nearly four times the amount of power that can be generated using the whole stream using AC-Direct technology can be generated using one fifth of the water using DC-power generation. Still, the resources are a drop in the bucket of full-project demands.

Hydroelectric development at Amelia Gardens would in the dry season only be capable of a nominal contribution, a truly demonstration project only and nothing more and arguably best associated with the amphitheatre or the education center to emphasize that the project is decorative and for instructional purposes. Winter plants might be viable in the winter, but without a properly calibrated purchase and sale interchange with the national power grid, the whole notion of independent power production is unlikely to be practical at all, irrespective of how the power is generated.

RAM PUMP / SWIM POND. The proposed location of a natural swimming pond and camping area 35 meters higher than the subjacent stream bed is easily serviced with a ram pump. A galvanized steel driveline with only 3.5 meters of vertical head is required to do the job. Ram pumps provide one gallon of delivered water for every ten gallons in the driveline. So, a decision is required on how the swim pond is to be managed. If it is continuously recharged, then an overflow will necessarily create an artificial drainage that is likely to sink in the soil before actually making it all the way back to the stream, particularly if it is a modest continuous pumping rate of 5 gpm or less. One significant drawback to ram pump usage is that the pumps are large, loud, and tend to be aesthetically objectionable. A more picturesque and quiet alternative to a ram pump is a water-wheel pump, though these in practice require larger flow rates to deliver comparable amounts of water. In the end it may be more practical to channel water by gravity to this site for a hypothetical pond.

TILAPIA / TROUT AQUACULTURE. The ephemeral stream drainage upon which the tilapia / trout ponds are shown does not have water in the dry season, and aquaculture requires fairly large amounts of water overturn due to the concentration of waste products among unnaturally concentrated fish. While tilapia are certainly adaptable to the tropics and are in fact hardy and can survive in oxygen-deprived settings ⁽⁵⁾, trout are not sustainable at Amelia Gardens. They require high dissolved oxygen concentrations ⁽⁶⁾, and oxygen solubility varies as an inverse function of temperature. For rainbow trout, which have an upper temperature limit of 75 degrees ⁽⁷⁾ but favor water temperature of around 55 degrees, you would need elevations of at least 4000 feet or so above sea level to provide a cool enough environment at southern Costa Rica’s tropical latitude of 8 degrees north of the Equator. In addition to the difficulties of having to introduce summertime water from somewhere, tilapia farming is relatively heavily regulated in Costa Rica, potentially complicating the process of permitting. Careful study is therefore recommended to determine the relativity of costs and benefits in aquacultural demonstration projects at Amelia Gardens.

WATER SLIDE. Similar to the limitations discussed above with aquaculture, the water slide would need water to make it work. Despite the fact that the ephemeral spring cannot provide year round water, since a water slide can reuse water, the water slide is considerably more plausible than aquaculture in terms of sustainable and viable resource development. It would appear from the master plan that the intention is to use water from the spring to convey water to the swimming pool via the waterslide. At times of the year when the spring is dry, it means that a water slide will necessarily depend on recycled water and the concomitant power investment to pump water and filtration to conform to Health Department requirements for a public water slide operation.

SWIM POND AND STORM WATER IMPOUNDMENTS. The collection of rain water in artificial impoundments is an option for expanding dry season water use capacity by creating storage in the form of ponds. However, the other side of the coin is that dropping water levels in relatively small impoundments may make breeding grounds for mosquitoes and frogs and may be aesthetically objectionable. The balance must be struck with the economics. Though the terrain is not optimal for ponds, they can be deployed in the watershed as conceived in the master plan, at a determinate capital and operating cost. It is unlikely that it will be economically possible to derive a hydroelectric contribution from this stored rainfall runoff. However, they could provide irrigation for permaculture installations. A concerted analysis of this would require greater information about the irrigation needs and locations in order to site collection ponds in parts of the watershed that will not only collect water but also provide for gravity flow to duty areas to minimize power requirements. There is a subtle environmental adverse consequence to the use of ponds that acts directly against stated project goals that must be mentioned. The creation of ponds eliminates all the original terrestrial plants busy consuming carbon dioxide and producing oxygen. Instead, the decay of organic material in the pond bottom produces two greenhouse gases: carbon dioxide and methane. Scientists have shown that as a consequence of greenhouse emissions from reservoirs, large-scale hydroelectric power generation has a larger carbon footprint than a comparable rate of power generation in oil-burning power plants ⁽⁸⁾. For this reason, large-scale hydroelectric power generation is not considered an environmentally sustainable despite being a renewable form of energy. Mini- and micro-hydro, which do not require reservoir creation, do not produce this effect and are therefore considered renewable resources and environmentally sustainable both.

PERMACULTURE. For routine lawn maintenance in Costa Rica, five mm of water per day is considered a baseline irrigation requirement. Since permaculture installations are likely to require more water per unit plant than grass but will have areas not planted as well, let us assume the rough criterion of 5 mm per day to serve as our assumed irrigation demand in the following analysis. Without trying to measure area on the design map provided, I will assume that there are three hectares facility wide that will want to be irrigated. Five mm of water per day over three hectares amounts to 40,000 gallons of water every 24-hour period. It is the equivalent of a continuous flow rate of 28 gpm. Unlike potable water consumption, which varies according to occupancy, irrigation requirements are constant, varying only as a function of the amount of area under permaculture and the water requirements of the plants being grown. Nevertheless, for the sake of argument, adequate irrigation of three hectares of land requires about the same as that required by 500 full-time residents using 75 gallons per day per person. The numbers indicate that at full occupancy, treated wastewater could theoretically provide 100% of needed irrigation rates.

Estimates for power demand were made on the basis of rough projections of appliance usage. Power demand was first calculated assuming the deployment of regular appliances and basic outfitting that would be expected of conventional American style homes. Air conditioning was not included in any of the projections.

Since the hydroelectric resources are on the order of less than one percent of the anticipated peak demand, the energetic requirements of Amelia Gardens must be satisfied by some means of power supply other than onsite hydroelectric, at least if dry season flow rates are the design criterion employed for power generation potential. While wind speeds were not measured and any possible contribution from wind power remains entirely speculative at this juncture, Amelia Gardens would be poorly served by a massive power generation capacity (whether wind, solar, or hydro) barring the ability to connect to the power grid and actually sell power generated in excess to the utility, irrespective of how the power is generated.

It was shown in a previous section of this report that dry season flow rate is nowhere near enough to satisfy energy demands through hydroelectric means. Other forms of onsite power generation include photovoltaic and concentrated solar power municipal-scale power farming. Wind power is another option that would on the surface appear worth additional scrutiny, since municipal-scale AC power generation can be achieved with wind turbines presuming average wind speeds exceed a baseline viability threshold of around 12 miles per hour ⁽⁹⁾. There is no doubt but that solar is a viable power option. However, before a single dollar is ever spent in development of any AC-power generation alternative, a guarantee must be worked out with ICE that guarantees purchase of excess power. In this manner a modestly-sized power generation capacity can be erected that feeds the power grid during periods of low occupancy, satisfies the facility with internal generation capacity during regular occupancy, and which depends on the grid to provide the extra power demand during periods of peak occupancy. This is essential since solar does not work at night and wind power is also subject to variations in atmospheric conditions.

The most practical first option to dramatically reduce the overall facility power demand is to require that each individual dwelling satisfy its own energy needs through the deployment of conventional roof-mounted solar power systems and corresponding inverter and battery set. This approach presupposes also the use of energy-efficient appliances, including DC-refrigeration, gas/electric hybrid dryers, and propane stoves and hot water heaters. Table 2 summarizes power demands for a well-anointed casita, excluding air conditioning, presuming energy efficient appliance use. Since many homeowners are likely to insist upon air conditioning, beefed up solar systems can be designed, based on house layout, and cooling requirements and the resulting power requirements, an analysis not included in this report.

Table 2. Power estimation of individual casita with efficient and hybrid appliances.

Table 2 reveals that a well-equipped house without air conditioning requires an inverter capable of a peak demand of 6.4 kw. The 18 kw of anticipated daily power consumption presupposes an array of 12 200 watt panels across an average eight hours of daily insolation. Such a system—presumably 24-Volt--would be best served with a battery bank of 12 1200 amp-hour 2-Volt cells. Such a system is likely to run around $25,000 installed.

For the cabins, it would appear reasonable to not include washing machine, dryer, dishwasher and to have fewer lights and fans so that the power demands might more closely resemble the more modest energy profile projected in Table 3. The lighter load would call for a system including: a 2.5 kw inverter, six 200-watt panels, four Rolls-Surette 512 amp-hour 6-V batteries, and a 60-amp charge controller for a total cabin power supply cost of around $12,000 installed.

Item	Qty	Watts	Hours/Day	Utility	Simultaneous	Total
Refrigerator	1	1500	24	0.5	1500	18000
Light Bulbs	15	75	5	0.6	1125	3375
Fans	4	65	18	0.8	260	3744
Washing Machine	1	1200	2	1	1200	2400
Dryer	1	1500	2	1	1500	3000
Dishwasher	1	1200	3	1	1200	3600
Miscellaneous	1	1500	1	1	1500	1500
Peak Usage					8285	35619

Item	Qty	Watts	Hours/Day	Utility	Simultaneous	Total
Refrigerator	1	120	24	0.5	120	1440
Light Bulbs	15	75	5	0.6	1125	3375
Fans	4	65	18	0.8	260	3744
Washing Machine	1	1200	2	1	1200	2400
Hybrid Dryer	1	1000	2	1	1000	2000
Dishwasher	1	1200	3	1	1200	3600
Miscellaneous	1	1500	1	1	1500	1500
Peak Usage					6405	18059

Item	Qty	Watts	Hours/Day	Utility	Simultaneous	Total
Refrigerator	1	120	24	0.5	120	1440
Light Bulbs	10	75	5	0.6	750	2250
Fans	3	65	18	0.8	195	2808
Miscellaneous	1	1500	1	1	1500	1500
Peak Usage					2565	7998