Alaska's Largest Solar-Thermal System - a Model for Multiresidential Use
In the summer of 2007, Willie Karidis came into my office at the electric utility where I worked with a challenging problem that, with time, would get worse.
Karidis was the executive director of the Denali Education Center (DEC), a nonprofit organization that houses guests visiting Denali National Park and Preserve during the peak tourist season in Alaska. DEC also houses employees who work there through the summer. This campus offers educational programs in Denali and supports park research. Besides housing and educating guests, DEC has meeting and eating facilities, offices, and an infrastructure for supplying heat and hot water for the residents. Karidis’s challenge was to control the soaring costs of energy at the DEC campus by finding appropriate ways to reduce energy use.
I set up a scoping meeting at DEC, which is located in a valley on the east side of the Parks Highway and south of a bridge spanning the Nenana River, which is part of the Denali National Park boundary. The meeting included staff members Karidis, Jodi Rodwell (who would soon replace Karidis as executive director), and Jill Boelsma. Our discussions focused on DEC’s high energy bills and on the potential savings to be realized by decreasing the energy load and installing a renewable-energy system or systems.
The cause of the high energy costs was an old infrastructure, built decades ago, that relied mostly on electricity, and some propane, to heat guest and staff housing and supply hot water during the tourist season. This is a seasonal-use (summer) facility. It is closed in the winter. The original owners did not insulate the cabins. When the facilities were first constructed, energy was cheap compared with today, and energy efficiency was not a priority. At the time of this project, most of the guest and staff housing is poorly insulated, if at all.
But energy costs have risen dramatically. During the summer, average temperatures in the area range from highs of around 60°F to lows of around 30°F. Some heat is necessary in the summer months, although temperatures occasionally reach the 80s.
Kick-Starting the Renewable-Energy Project
DEC had virtually no extra money for large capital improvements. However, money did become available in 2008 through the state of Alaska’s Renewable-Energy Fund Grant Application Recommendation Program for the purpose of financing, among other things, a feasibility study and the construction of renewable-energy projects that met certain criteria. The electric utility that serves DEC could apply for such financing.
After a number of scoping meetings, the group tentatively agreed that a solar water-heating system (SWHS) for the DEC campus was the best renewable-energy option. So with the approval of the utility, I prepared grant proposals, first for a feasibility study and then for the construction of the solar-thermal system. The goal of the project was to create a renewable-energy resource that could offset an existing fossil-fuel-generated energy source and contribute to the development of a renewable-energy industry in the state. The funding for the feasibility study—and later the construction project—was approved by the state legislature.
The feasibility study allowed us to rule out wind power—there are no sufficiently high wind speeds in the valley where DEC is located. We considered a hydrokinetic river turbine, since the campus backs up against the Nenana River. Costs, maintenance requirements, seasonal removal of the turbine (since the river freezes over every winter), and other issues moved this option off the table. Given the Alaska summer sun and the seasonal nature of the business, we decided that solar power—either PV or thermal—was our best option. We wanted to offset a primary baseload energy load (that is, heating water), and we knew that water heated by the sun in the summer could be stored less expensively than if we used a system that required PV and batteries. Most of the campus—particularly the residential portion—was mothballed during the winter. Also, at the time, the grant required a renewable-energy system, so an air-source heat pump water heater coupled with a PV array would not meet the requirement.
By offsetting electricity use through a solar-thermal system, we could reduce operating costs to improve DEC’s financial balance sheet, and lower the load on the organization’s facilities. A lower load could prompt a change in the utility rate on the DEC account by eliminating a punishing demand charge on its bill that is normally levied to high-load customers. The SWHS would become an excellent example of load management. In addition, since the public is keenly aware of global warming and of businesses that practice sustainability, DEC was demonstrating its commitment to preserving the environment by using renewable energy. In Alaska’s hospitality industry, customers are more likely to spend their money with businesses that are eco-friendly, knowing that they share the same values when it comes to the environment. So the project demonstrated something that could be useful to the industry and was viewed as a public benefit by the state.
Determining Energy Loads, Demand, and Solar Capacity
We analyzed DEC’s entire energy load, identified and separated out water- heating loads, and estimated current and future water-heating demands. We then sized and designed the SWHS for construction.
The DEC campus covers about 10 acres (see Figure 1). There are 18 buildings in the northern half of the campus. There are several employee cabins scattered throughout the southern half, but these cabins have no plumbing and were not included in the analysis. The 18 primary buildings consist of 13 housing units, a laundry building (which also houses the water distribution system), a community dining building, several staff office buildings, and an employee kitchen-and-laundry building. All buildings have plumbing and water and space heating. The housing, laundry, and offices have electric water heaters. The community dining building, which includes a large kitchen space, uses propane to heat water.
Domestic-use water is supplied by a well that was drilled on the southeast edge of the campus. It is piped to the laundry building, where it is diverted into two primary 1½-inch polyethylene distribution lines, one extending to the employee kitchen-and-laundry building and the other to an office building with feeder spur lines to the guest housing units.
Because we had only a few months to prepare the funding proposal, and because we had to do it outside the seasonal peak-load months, when only a skeleton staff was working, we were not able to log and collect real-time energy use data. Instead, we analyzed hot-water use by comparing past electricity and propane use records to a conventional-population average use—a method that the tourism industry uses. The population method relies on assumed daily hot-water use per person, without considering time of use or specific demographics. It estimates energy consumption by multiplying average assumed daily hot-water use per person by the average number of individuals at the facility. This quantity is then multiplied by the quantity of energy required to heat a unit of water. The accuracy of this method depends on correct population data and on realistic assumptions concerning usage.
The historical energy method described below is based upon determining the average energy consumption of the facility and extrapolating what fraction of this total energy use is devoted to heating water. We compared findings using this method to our findings using the population method. Neither of these methods represents an exact science, and we wanted to learn if the findings were similar.
Energy utilities regularly conduct load profile surveys and projected load forecasts to anticipate necessary power generation capacity. In our analysis, we chose to focus on energy use and not on load profile, since we assumed that our renewable-energy system—solar thermal—would augment, but would not entirely replace, fossil-fuel-derived energy.
Population Method of Assessing Energy Use
The DEC staff estimated that the campus population averaged about 64 people per day during the operating season, which runs from mid-May to mid-September—about 120 days, or four months. At the time, the hospitality industry estimated that activities such as showers, laundry, and cooking require about 20 gallons of 122°F water per person per day. Using these estimates, we assumed that DEC would use about 1,280 gallons of heated water every day.
The math was relatively straightforward. It requires 1 Btu to raise 1lb of water 1°F. Assuming that the cold-water supply (from the well) is 35°F, and that hot water is dispensed at 122°F, the water temperature must be raised by 87°F to meet the desired temperature. A gallon of water weighs 8.3 lb, so each gallon of water requires 722 Btu (87°F x 8.3 lb) to raise its temperature to 122°F. If DEC consumed an average of 1,280 gallons per day, the water-heating demand would be 924,200 Btu per day (1,280 gallons x 8.3 lb per gallon x 87°F). This equals about 270 kWh per day, since 1 kWh equals 3,412 Btu.
Historical Utility Usage Method of Assessing Energy Use
We compared these numbers with our findings using the historical utility usage method. From past electric usage statements, we estimated daily kWh average use over previous operating seasons. We calculated the daily kWh average use for the 120 days of the operating tourist season to be 406 kWh per day. Each of the 13 housing units used electricity for both water and space heating, and we all intuitively agreed that these loads were responsible for most of the electricity use. We attributed 90% of the total campus electricity consumption to the housing units. Thus 406 (total seasonal average) multiplied by 0.9 (90% of total load) equaled an average daily use of 368 kWh per day. We then assumed a 50/50 split between water and space heating—it was an educated guess based on users’ experience, bringing average consumption for hot water to 183 kWh per day.
Total historical propane use for the buildings that heated water with this fuel (we used the previous-year consumption totals) was 127 gallons for an office and 647 gallons for the community dining hall and kitchen. We estimated that 95% of that use in the office and 50% of that use in the dining hall and kitchen went to heating water. This gave us an estimated propane consumption of 445 gallons for heating water during the previous year. This is the energy equivalent of 9,030 kWh, using a thermal conversion efficiency of about 75%. This use added 60 kWh to the total average daily consumption. When combined with our electricity use estimates, it brought average energy consumption for hot water to about 243 kWh per day.
The estimated energy demand from the population method was 270 kWh per day, and the estimated energy demand from the historical method was 243 kWh per day. We averaged the two findings to arrive at 258 kWh per day as the current water-heating demand. We then anticipated an increase in energy use at 5% per year over five years for a total of 330 kWh, or 1,125,960 Btu, per day, average at the end of five years. We estimated that 50% of this daily average energy use could be offset by solar water heating over the heating season. This estimate was based on the assumption that the number of users, and therefore energy use, would increase. We felt that not only would the solar water-heating system attract more customers, but also that it might open up the “shoulder” weeks, allowing the campus to open earlier and close later. That has not happened, and to my knowledge, the growth in customers during the season has been about flat, and energy use has not increased.
Without the Energy Center building and the equipment it houses, the array would simply be a stand-alone piece of art.
We analyzed solar capacity using the Polysun solar simulation software program and compared it with real-time solar data (W/m2) collected at the headquarters of Denali National Park and provided by the National Park Service. Long-term solar insolation data were also available for Fairbanks, but while these data were reliable to a degree, it was more difficult to interpret how the resource might vary at DEC, because the campus is located in the mountains, about 130 miles south of Fairbanks. Differences in cloud cover were also a concern, since cloud cover directly affects the resource.
The Polysun profile was somewhat lower than the historical solar conditions of Fairbanks and the data collected in the national park. We felt that the Polysun solar insolation profile was on the conservative side of accurate, so we used it for the capacity analysis.
Siting the SWHS
Virtually all of the DEC campus is forested with spruce trees up to 40 feet high. Concerns about aesthetics, environmental impact, easy access, and easy maintenance all influenced our decision to install a ground-mounted centralized array. We also thought that the south-facing array might help the public to understand how to optimize systems that draw on solar energy. The ground-mounted array would serve each of the designated buildings with a flexible insulated buried pipeline. The site did not have any restrictions that would significantly decrease its performance. The hills to the south were low enough that if they blocked the sunlight in December and January, it would be at a time when solar energy was not needed. During the solar-heating season, local weather conditions (foggy mornings or cloudy afternoons) could affect the optimal orientation of the array, but this was the case for just about any solar collection site.
A natural clearing on the south side of the campus near an access road and between the laundry facilities and the first guest cabin provided an excellent site for the solar array. It was adjacent to the proposed DHW circulation loop, close to the existing water source, and accessible for educational purposes without intruding on the campus setting.
The array would consist of 36 south-facing 4-foot x 10-foot Heliodyne Gobi 410 collectors connected in six parallel strings and installed on a fixed ground-mounted structure. We selected flat-plate collectors over evacuated-tube collectors for their durability and ease of maintenance. From the solar array, an indirect closed-loop system of propylene glycol would transfer heat to a 2,800-gallon insulated polyethylene heat-exchange tank by way of a set of copper coils seated in the tank. The tank would also provide thermal-energy storage. When the temperature of this fluid at the collectors exceeded the temperature of the tank, a controller would activate a circulation pump.
A second set of copper coils connected to the domestic-water distribution system, and also located in the heat exchange water tank, would move the now-heated water to the individual water-heating storage tanks located in each of the buildings to be served. The district-heating loop would consist of 565 feet of 1½-inch insulated PEX pipe running to the dining hall. From the dining hall, an automated booster pump assembly would increase the water pressure, when necessary, to the remaining 754 feet of 1¼-inch insulated PEX pipe in the loop. Unused hot water would return to the central water storage heat exchange tank, and the cycle would continue.
A minimal circulation flow of 1–2 gallons per minute (gpm) in the district-heating loop would ensure that water in the loop did not stagnate and cool excessively while minimizing heat loss within the piping system. (It is not necessary to supply hot water 24 hours a day, unless the system is retrofitted in the future to supply space heating to replace or augment the existing electric heat. In that case, other control strategies may be useful.)
All buildings would be serviced directly from the primary district loop except the employee kitchen and laundry, which would be on a 125-foot spur of 1-inch insulated pipe, and the guest laundry building, which would be on a 126-foot spur of 1-inch insulated pipe. The potential existed for water to sit and cool down in the two spur lines, but it would amount to only about 5.4 gallons in each line.
Each hot-water station would be connected to the distribution loop by way of a tee. Each station would consist of a hot-water storage tank with an internal electric or propane heater and a tempering valve. The water from the distribution loop would feed into the cold-water inlet of the station tank, and the internal heater would be activated only if the solar-thermal system could not keep pace with demand. The tempering valve would regulate the tank outlet temperature to a safe level. The design assumed that we would utilize the existing hot-water stations at the site. See Figures 2 and 3 for schematics of the whole system design.
The individual hot-water stations meet hot-water demand. Demand consists of showers, laundry, and cooking and is assumed for the purpose of this model to be 1,200 gallons per day for the period mid-May to mid-September. This number was arrived at in the load evaluation phase of the project. Based on existing fixtures, we calculated the maximum instantaneous demand to be 70 gpm. System sizing was based on the expectation that actual demand would not exceed 45 gpm. Several conservation measures were incorporated to reduce overall demand and mitigate peak demand. These measures included installing sensor heads for faucets, installing low-flow showerheads, using cold water for laundry, and installing insulating blankets on the existing water heaters.
Unfortunately, the solar system was designed for the old plumbing. This was an oversight and a lesson to be learned. However, the funding proposal did not allow for costs other than the cost of the renewable-energy resources. The water-heating system was sized to allow for anticipated growth in demand (as I explained above), but it was not sized to allow for upgrading the plumbing. The solar simulation model indicated that this system design would yield a solar fraction exceeding 70%. This model takes into account weather and seasonal variations, collector efficiency, thermal losses, demand variation, heat exchanger efficiency, and so on.
In the late spring of 2009, construction on the SWHS commenced. The site was relatively clear, but a few trees were removed. Materials were staged around the campus. Holes were drilled to set pilings for both the energy center (where the heat exchange water tank, the interconnection of the closed solar loop and distribution infrastructure, the control panel, and the circulating pumps were housed) and the solar array. Ditches were cut to lay the insulated plastic distribution pipe. Existing water-heating storage tanks in housing units and other buildings were retrofitted with new plumbing to connect into the new distribution lines.
Construction then started on the array frame, the mounting of the collectors, the necessary plumbing and electrical, and the energy center. By August, the SWHS was completed, tested, and commissioned. Operation time for the SWHS in 2009 was only a few weeks before the campus shut down for the winter. The whole system was then drained, and plumbing was blown with air.
The total capital cost for the SWHS was about $210,000. This did not include the feasibility study or the many hours of contributed labor.
At the end of September 2013, the DEC SWHS had been in operation for three full seasons. Preliminary projection for annual energy displacement estimated about 36,000 kWh and about $9,600 saved per year (at cost value in 2008). Savings in energy and cost have been realized, but in reality those projections have not been met—at least, not yet.
The solar collection array, its plumbing, and the system control work well. The array has had few serious problems. A pressure valve at the top of the array needed to be replaced. A circulating pump was also replaced. Several of the dozen monitoring sensors needed to be repositioned or replaced.
The lower-than-expected performance is due to many factors, some of which we could have anticipated and avoided—by improving the design or by assessing other case studies more thoroughly. The SWHS is a whole system. Simply retrofitting renewable technology in the front end does not guarantee that energy and money will be saved. It is also necessary to research the distribution system and to design the retrofit accordingly. But this can be expensive, and funding was a limiting factor in planning. So we worked with what we had and fixed the problems afterwards as operating money was made available.
The SWHS is a work in progress. It is important to make the buildings as energy efficient as possible to reduce unnecessary load requirements. The system requires deliberate monitoring and regular maintenance—not fixing, just attention. Because the load is irregularly dispersed, the SWHS sometimes overheats the fluid in the collectors. Demand becomes stagnant during the day when guests are gone and the system needs an outlet. Providing a larger-capacity storage tank—or a heat dump, such as a hot tub—might make the overheating less of a problem.
Because the existing water heaters are old and there are minerals in the water, scaling occasionally blocks the distribution lines and the pumps. It was necessary to upgrade the existing demand side of the system in order for it to operate effectively. After construction, all the existing individual water storage tanks in the buildings were replaced in an attempt to minimize scaling. Shower valves were also retrofitted with new hot-water temperature control valves.
Dind out more about and perhaps visit the Denali Education Center.
Distributed Generation Application
Despite these obstacles, the SWHS works well. Lessons learned are ongoing. I believe the project, where a cluster of housing units are served by a single SWHS that distributes heat, can serve as a model, especially in new construction, where the demand and supply side can be more accurately calculated. This project has the potential to save much more energy and much more money.
Enter your comments in the box below:
(Please note that all comments are subject to review prior to posting.)
While we will do our best to monitor all comments and blog posts for accuracy and relevancy, Home Energy is not responsible for content posted by our readers or third parties. Home Energy reserves the right to edit or remove comments or blog posts that do not meet our community guidelines.