Final Project Report

Solar Hot Water

Steve Phillips Jeremiah Robinson Josh Weidler Messiah College ENGR 491 Dr. Donald Pratt May 13, 2002

Abstract Our project group, the Zapatismo Musketeers, under the supervision of our advisor, Professor Carl Erikson, designed and prototyped an active solar hot water heating system intended to operate in conjunction with typical existing domestic hot water systems. After reviewing literature to establish the state of the art for active solar water heating, we developed some new innovations to improve the performance of our system. The design features a flat-plate solar collector with improved heat transfer characteristics and a microprocessor based controller, which allows for more complete system monitoring than a typical controller. These improvements to typical solar hot water systems will increase the efficiency and desirability of such a system for the average American homeowner. After completing the system design, we built first prototypes of both the solar collector and control system and performed preliminary testing of each.

Table of Contents Acknowledgments ..............................................................................................................1 1. Introduction....................................................................................................................2 1.1 Description...............................................................................................................2 1.2 Literature Review.....................................................................................................6 1.3 Solution ..................................................................................................................10 2. Design Process ..............................................................................................................14 2.1 Solar Collector Design...........................................................................................14 2.2 Control System Design ..........................................................................................17 3. Implementation ............................................................................................................19 3.1 Collector Construction ...........................................................................................19 3.2 Control System Construction.................................................................................22 3.3 Operation................................................................................................................23 4. Project Management....................................................................................................27 5. Budget ...........................................................................................................................28 6. Conclusions...................................................................................................................30 7. Future Work.................................................................................................................35 Bibliography .....................................................................................................................37 Appendix...........................................................................................................................38

Acknowledgments We would like to thank our advisor, Professor Carl Erikson, for his guidance and also for allowing us to use his farm as a testing location for our project. We also want to thank Energy Kinetics, Inc. of Lebanon, New Jersey for providing us with a hot water tank, circulator, and other components necessary for testing our system. Finally, we extend a special thanks to Dr. Don Pratt, Earl Swope, and John Meyer for advice and assistance with design and construction.

1. Introduction 1.1. Description God calls us to be conscientious stewards of his creation, and our motivation for developing this product comes in response to this calling. While our project is specific in its purpose and implementation, it addresses large and important issues. These issues include energy availability, environmental responsibility, and waste. The United States rested since the energy crisis of the 1970’s in the confidence that our resources would continue to provide for our needs well into the future. Now however, it is apparent that these needs are not always being met. The energy crisis in California has forced the U.S. to look again at our energy resources and consider tapping into new ones. By exploring the natural and sustainable energy found in the sun we hope to offer a small contribution to the solution of the energy crisis at hand. Another major issue is the environment. Almost all of the United States’ energy comes from non-renewable resources, such as oil and coal. Most of these resources are not only being depleted but they also harm our land when taken for fuel and harm our atmosphere when burned. The removal of these assets from the ground can deeply scar the habitats of animals and even people in these areas. This can be seen with the recent debates over removing oil from the forests and habitats in Alaskan national wildlife preserves. As resources become scarce, more habitats will be destroyed and some animals will become endangered or even extinct. By improving sustainable technology we respect and help to protect these precious lands. We need to examine our energy use and waste not only because of what it means to us, but also because of the effect it has on people and lands around the world. As the recent terrorist events have shown, many people in other countries have a severe hatred of our country. Our first response to these acts, before any retaliatory actions, should be serious introspection and examination of our policies, including our energy policies. Some people’s resentment of the United States may stem from our sometimes heavyhanded dealings with oil producing nations. By developing sustainable energy technologies, our reliance on crude oil can be reduced, changing the face of American foreign policy. This benefit of an increased dependence on solar energy should not be overlooked. By using state of the art technology to tap into an unlimited and harmless

source of energy, we hope to show that we can live a comfortable life in a more responsible way, as we respect God’s creation and our world neighbors. Our purpose for this project is to reduce consumption of fossil fuels by using alternative technologies in the heating of domestic hot water. In order to define the goals of our project and to provide a means of evaluating the project’s success, we proposed the following objectives: • • • • • •

System freeze protected to –10°F and able to withstand temperatures up to 110°F, using only automatic controls (no manual maintenance). Save greater than 40% of yearly hot water heating costs. Comply with all government plumbing, building, and wiring codes. Meet a $1000 cost target for installed unit. Fully automate system to require no more than a yearly maintenance check. Configure control system to allow for interface with different existing hot water systems.

Taking a general overview of our objectives, there are several possible ways to meet them. Recent advances in current power plant technology could allow a power plant to provide for our objectives. Unfortunately, this type of investigation would be beyond the scope of a senior project, and most likely not the most environmentally sound option. Other options for individual homeowners may include geo-thermal energy, biological methane energy, and photovoltaics. In the case of geo-thermal or bio-methane, these processes may be sustainable and environmentally suitable. However, they are not as universally available as sun energy. Photovoltaics should be considered. At first, it may seem wasteful to convert solar energy to electrical energy and then to heat as opposed to converting sun energy directly to heat. However, electrical energy could also be more universally applied to other household systems, even providing the energy necessary for the control system and circulator pump required for hot water heating. Another way to save up to 40% of the water-heating bill would be to take a different approach to the problem and deal with the appliances that actually use this energy. Water efficient showerheads, dishwashers, and clothes washers could, in many cases, save 40% on water heating. Low technology solutions should be carefully investigated and used wherever possible in our design. In most cases, for small-scale applications, the more low

technology, the more economical the solution. The most daunting challenge when considering these options would be in our objectives of automation and reliability. Our project is going to be designed for use by an American audience, and therefore must have the convenience to which they are already accustomed. There are many different styles of solar water heaters currently on the market. Some are direct, which means that potable water passes directly through the solar collector. Others are indirect, which means that the heated fluid and the potable water are kept separate and heat is transferred by means of a heat exchanger. Indirect systems can have antifreeze in pipes that are exposed to freezing temperatures, because the potable water never mixes with the toxic antifreeze fluid in the closed loop of the heating system. Direct systems must be drained (either manually or automatically) in winter to prevent the bursting of pipes due to freezing conditions. The collector itself offers several possibilities for design. In widespread use today, there are batch collectors, flat-plate collectors, evacuated tube collectors, and parabolic collectors of various designs. Batch collectors consist simply of a large tank placed in a sunny area. The tank may be hidden behind a glass glazing in an insulated box to improve heat capture, with a coating on the surface of the tank to allow heat to pass into the fluid through the walls of the tank. The benefits to the batch collector are that it is extremely simple to build and can be made from inexpensive materials, and that it does not require an additional storage tank. The disadvantages are the structural concerns with the weight of the fluid if the collector is placed on the roof, and the inefficiency of heat transfer to such a large volume. Another type of solar collector is the flat-plate. It consists of pipes running over an absorptive plate through a thin insulated box with a large surface area and a glass glazing covering the surface. These collectors improve the efficiency of the heat transfer by reducing the volume of water inside the collector, and solve problems of weight on a roof. They also look more stylish on a home than the thicker batch collector box. However, flat-plate collectors are more difficult to construct and more expensive because of materials, and require a storage tank to store the heated water. Evacuated tube and parabolic collectors lie on the high end of the technology spectrum for solar hot water heating. Evacuated tube collectors are similar to flat-plate collectors, with the air removed from the collector box, allowing for maximum

heat transfer. Parabolic collectors consist of a thick tube of fluid running inside a collector box, with a reflective plate shaped in a parabolic shape, reflecting maximum sunlight directly at the tube. Both of these options are extremely efficient. However, they are more difficult to construct and require materials that are more expensive. They also pose potential problems, like the possibility of a leak in a vacuum-sealed box or the generation of thousand-degree temperatures in a parabolic collector. Like the collector, the requirements of the control system can be met by employing any one of several feasible options. The options include using no control system, a simple differential switch with thermocouples, or a microprocessor-based controller. Control systems, ignoring labor costs, do not vary significantly in price; all cost less than $25. Therefore, other factors much more heavily weigh into the decision. The first option, eliminating the control system altogether greatly simplifies the design and eases installation. In this scenario, the pumps run continuously, the fluid flowing through the collector day and night. This would simplify freeze protection and heat protection, as fluid flowing continuously would have no opportunity to sit in the collector and freeze or become too hot. A serious disadvantage, though, is that the heat transferred to the tank during the day would be radiated out at night, particularly in the winter. Because normal hot water needs require the use of most water in the mornings for showers, loss of heat at night would be detrimental. A differential switch operates by turning pumps on when the water in the collector is a certain temperature above the water in the solar hot water tank and turning them off when the temperatures are equalized. The differential switch typically uses thermocouples to sense the pertinent temperatures. These thermocouples can attain temperatures accurately, but require quite a bit of design work and calibration. After the initial work required setting up the differential switch, this control scheme is very simple and has been proven to be reliable. However, this type of controller does not offer particularly high accuracy and the simplicity precludes the use of complicated algorithms that might improve efficiency. They also provide no options for freeze or heat protection, save the use of a more concentrated antifreeze solution. The final option, a microprocessor-based controller, provides more efficiency but is more complicated and requires more technological expertise for the designer. It allows

for complicated algorithms and high levels of accuracy with high-tech temperature sensors, squeezing every last BTU out of the collector. It also opens up the possibility of taking over control from the existing hot water system, fully integrating it with the solar system and saving some additional run time on the boiler or the tank burner. Furthermore, it allows for freeze and heat protection with the microprocessor’s ability to work with multiple sensors, each with its own function. The significant disadvantage to the microprocessor system is its high level of technology and difficulty of programming. To make the design feasible to put into production and offset the initial design labor costs, the numbers of systems produced must be high. 1.2. Literature Review

Closed Loop System Closed loop systems use either antifreeze (non-toxic propylene glycol) or water as the medium. An antifreeze system is the only one that allows for freeze protection. In order to transfer the heat from the collector (which is filled with antifreeze) to potable water, a heat exchanger becomes necessary. Closed loop antifreeze systems, however, are the most complicated. They contain an expansion tank to allow the antifreeze to expand and contract with changes in temperature. They also must have a pressure relief valve to protect against excessive pressures in the closed loop. In order to prevent gravity flow of hot water from the tank out to the collector, which would cause heat to be lost at night, a spring-loaded check valve is used. An air vent is needed to help remove air from the closed loop, and a pressure gauge is inserted to tell if your system is still charged. A couple of temperature gauges are a good idea in any system so you can tell how well it’s working. The system also needs a charging pump. This can be utilized with two boiler drains and a shutoff valve. The pump charges the loop with the antifreeze solution, expels all the air upon startup, and maintains this condition during operation. Despite their complexity, these systems have a high degree of reliability and are well understood by heating contractors. There is a downside, however, to the closed loop antifreeze system. In the heat of the summer, when the system finishes heating the tank to the set temperature, the solution

just sits in the collector and stagnates. Under this condition, temperatures can climb to as high as 400°F. At high stagnation temperatures, air pockets can form and propylene glycol can begin to break down and come out of solution. An inhibitor can help, but they are expensive and don’t usually work with off-the shelf piping components. General Collector Information For applications requiring water no hotter than 140°F, flat plate collectors are the way to go. One such design consists of copper tubes sitting on a flat absorber plate. Usually this means a series of parallel tubes connected at each end by pipes called the inlet and outlet manifolds. This assembly is covered with low-iron, tempered glass as the glazing. The most efficient collector design maximizes solar heat gain, minimizes heat losses, and provides for the most efficient heat transfer from absorber plate to the pipes. Temperatures as high as 250°F are possible at times, but this should be avoided in a domestic hot water system. The absorber surface is usually made of black chrome. This is far more efficient than just a black painted piece of sheet metal. Although a painted black surface is efficient at absorbing solar radiation and converting it to heat, it re-radiates infrared heat back out. These losses reduce collector efficiency. A highly polished chrome surface would lose the least infrared heat energy. However, not being black, it wouldn’t absorb anything. Black chrome is the best of both worlds, with high absorptivity and low emmissivity. It’s universally available too, with only a marginal additional cost. It is especially useful in cold climates, where radiant heat loss is a serious problem. Solar hot water systems consist of five basic components. They are the collector, the absorber, the storage, the distribution system, and the regulation. The collector, also known as the glazing, is the part that lets the sunlight into the system. The absorber is the recipient of the solar radiation, which is transferred into thermal energy. It is usually black, as this color has the highest absorptivity. Storage is the medium where the heat energy is stored, usually water or another dense liquid, but possibly another dense material such as concrete. The distribution system carries the stored energy to where it is to be used. This component may require a pump. The final component is regulation.

This component is usually implemented by an electrical control system, or may have been mechanically incorporated into the design. Flat Plate Collectors One in every fifty homes in the U.S. has some type of solar collector installed and the majority of these are flat plate collectors. Flat plate collectors, while not very efficient at heating a house, can be very effective at heating the hot water necessary for household usage. A maintenance free system, though, has not yet been perfected. The simplest and roughest type of a flat plate collector is known as the trickle type collector. This low-tech system includes an inlet pipe with several holes where water trickles out and onto a black painted corrugated aluminum sheet. After flowing down the sheet the water is sent through a return channel. A more complex and reliable system is the tube in plate type. This has several advantages over the trickle type. First the water is not exposed to airborne contamination, also there is no evaporation and the fluid is heated by a larger heating surface. The efficiency of the collector is determined by its thermal conductivity, its losses to surroundings, its ability to absorb energy, and the path of the fluid through the plate. The more constant we can hold the temperature of the plate, the more efficient the system. Choice of material for the absorber is one of the most important issues in design. There are trade-offs between thermal characteristics, cost, and material joining. The coefficient of thermal expansion is extremely important for bonding applications. It is most effective to use either copper or aluminum. Another important factor in the design is the tube configuration and spacing. Tube spacing in parallel gives a good temperature distribution, but is more costly. There is not much of a pressure drop but it is not reliable if it needs to be drained. Series or serpentine configuration has a lower efficiency due to non-uniform temperature distributions and a higher-pressure drop, but is more easily drained. Many people have attempted to find ways to make a more cost efficient solar collector. One way that has proven very effective is to use integrated flat plate solar collectors. That is to install the collectors into the structure of a home during

construction. Another cost effective method is the use of plastic in the design. When using plastic however, the limitations of durability must be considered. The following rough estimates for daily insolation have been observed for flat plate collectors:

300 BTU/sq ft -

sunny day

50 BTU/sq ft -

cloudy day

Glazing/Collector When operating at a temperature difference between the collector plate and the ambient air temperature of greater than 20°F, a glazing becomes necessary in order to trap heat. The glazing material is usually glass, but plastic can be used if careful attention is paid to possible softening and warping due to heat. In the winter it is advisable to use a double paned glazing to better insulate the collector. This double-glazing, however, decreases the optical efficiency of the unit, not allowing as much light to enter the collector. A low iron content glass is typically used because it has a high optical efficiency. This glazing is usually tempered for strength. As the collector is mounted outside, it needs to be strong enough to weather the elements. Piping/Installation Because the water heated by solar energy in a flat plate collector can reach very high temperatures, an anti-scald mix valve becomes necessary. A mix valve regulates the temperature of the water that is allowed to flow to a faucet. If the water heated by the collector rises above 120°F, the mix valve will temper that water by mixing it with cold water. This device is necessary to protect the users from excessive temperatures, which may become present in the collector. Another required safety device is a pressure and temperature relief valve (PTRV). This device must be installed at the hot water outlet of the heater in case temperatures or pressures become really high. You will find one of these valves installed on any hot water tank as well. It’s required by code, but only operates in an emergency, and usually needs to be replaced if it opens. Some local codes require some extra separation between the heat-transfer fluid and the potable water in closed-loop systems. This may rule out single-wall heat exchangers. Double walled heat exchangers are available for solar hot water

applications. Most local codes, though, allow you to use a single walled heat exchanger when the fluid is non-toxic. You usually put it inside the water heater for maximum transfer efficiency, but a separate unit outside the tank can be used to house the heat exchanger if necessary. Be sure that it is accessible and well insulated. The type of heat exchanger can have a huge impact on the overall efficiency of the solar system. They can range in efficiency from less than 40% to about 90%. If you have two tanks, the system must provide for flow from one tank to the collector and back. Small (10 to 30 watt) circulating pumps provide the necessary flow. The direct pump system uses an electric circulating pump to move heat from the collector to the storage tank. This means that you don’t have to put the collector below the tank, as you would for thermosiphon flow. The pump can move heated fluid from the collectors on the roof to a storage tank in the basement. Good sense still calls for minimal length of pipe run for efficiency, though. Solar hot water system pumps can be powered by photovoltaic cells. These pumps need to run on the DC power provided by the PV panels. Though simple, they tend to be reliable and efficient, providing hot water when other systems have been shut down by power outages. The most efficient orientation year-round for the collector is facing south, tilted at an angle about the same as the latitude of the site. Make sure to aim it towards true south, rather than compass south. Adding 15º to the collector tilt in the winter and subtracting it in the summer maximizes efficiency. For cost savings and aesthetic reasons, however, most people lay them flat against pitched roofs. This does not typically affect efficiency in a dramatic fashion. The exact tilt of a collector is not crucial; a 10º variation to suit a roof's pitch usually makes no appreciable difference.

1.3. Solution After conducting detailed research in the area of solar hot water heating and traditional hot water heating, we designed and built a solar hot water heating system that can be integrated with a homeowner’s current domestic hot water system. This system would conserve energy and save money for the homeowner, while not compromising the effectiveness or reliability of a traditional system. While not completely eliminating the

need for electricity, gas, or heating oil, this technology takes a positive step in the transition from fossil fuel power to more sustainable energy sources. Our system provides meaningful advances in the efficiency and desirability of the current industry standards for active solar water heaters. Although there exists a wide variation in traditional domestic hot water schemes, our system can interface with nearly any scheme. Some homes have an electric, gas, or oil fired hot water tank separate from their home heating system. Others use their boiler, which provides heat for the house, as a source for domestic hot water. These systems can utilize either a tankless coil boiler or an indirect system with a storage tank. Each of these schemes could benefit from an integrated solar collector to provide extra (free) energy to preheat the incoming hot water. Our system may be installed into a house with any of these domestic hot water systems. We explored and discussed the different styles of solar water heaters currently on the market. Indirect systems can have antifreeze in pipes that are exposed to freezing temperatures because the potable water never mixes with the toxic antifreeze fluid in the closed loop of the heating system. In contrast, direct systems must be drained (either manually or automatically) in winter to prevent the bursting of pipes due to freezing conditions. Because direct systems require draining and yearly maintenance, we designed a closed loop, indirect system. The water/antifreeze solution that is in the solar heater loop will never mix with the potable water. We explored different heat exchanger options to make sure that the antifreeze never mixes with potable water to avoid obvious health hazards. A closed loop, indirect solar water heating system has certain challenges, but accomplishes our goals most effectively. The basic elements of the closed loop, indirect solar water heating system are a solar collector, a tank, two pumps, a heat exchanger, piping, and a control system. The solar collector can be combined with a wide variety of systems, so we will work with only one collector design. We designed and sized the collector, through heat transfer analysis, to meet our stated objectives. While the market contains many different types of solar collectors, we designed and constructed a flat plate collector. We chose this type to avoid the expense and complications of the more high-tech versions and the bulkiness and inefficiency of the batch heaters.

Figure 1

Our system requires a hot water storage tank. This solar tank is necessary even if a conventional tank is already present. The collector will capture excess energy during the day, which will need to be stored overnight while the collector will obviously not be collecting any energy from the sun, for use the following morning. On an average night, the moon, stars, constellations, black holes, etc… do not provide enough energy to appreciably heat the water. We determined the optimum storage tank size for use with the solar collector and water temperature needs and built accordingly. Two pumps circulate the water/antifreeze solution heated in the solar collector in one loop and the water heated in the tank in the other loop, both through the heat exchanger. Both of these pumps must run in order for fluid to flow through both sides of the heat exchanger and for heat to be transferred. Our pumps run on AC power and are commonly used in the heating industry. The control system operates the pumps. Along with operating the solar system components, the control system interfaces with the existing system and actually overrides the existing hot water controls. Our control unit makes a major improvement over industry standards and makes this system adaptable and therefore more universally marketable than current solar heating technologies. Our controls system uses a Motorola 68HC11 microprocessor. It monitors four different system temperatures with digital integrated circuit sensors. Based on the information reported by these sensors, the processor controls the system components. A description of the control strategy can be found in the design section of the Engineering

Design Report. An exciting aspect of the control design is the interface with the existing hot water system. Our control unit becomes the primary control, deciding when to turn the existing system on and off. Our system greatly reduces the amount of energy that the homeowner must purchase to make hot water. During the hotter months of the year, this system will generally provide for all of the home’s hot water needs. Consequently, our product should help ease the transition to sustainable energy sources.

1. Introduction 2. Design Process In designing the complete solar hot water system, we broke the project into two major components: the collector and the control system. Each system component provided its own challenges. After completing thorough analysis and design in the fall semester (see EDR), we began to do some preliminary testing to determine the best way to implement our design in a prototype. These initial prototyping experiences led us to some design changes. The design process that we completed in the spring semester followed from the work we did in the fall and is discussed here. 2.1. Solar Collector Design We completed design work on three areas of the collector: the glazing, the insulation, and the absorber plate. For the glazing we had hoped to obtain a glass with a thickness of 1/4” to provide extra insulation in cold weather. Our final prototype used a 1/8” thick piece of low iron tempered glass. This glass was provided to us at no charge, and to buy our own glass would have put the project over budget. The Celotex insulation called for by the original design was available, and this aspect of the design remained the same. The absorber plate was the most extensive design consideration for the collector and experienced the most design changes. The two major changes necessary for prototyping were the riser/foil plate interface and the material of the system. Several small preliminary prototypes were constructed and tested to determine whether they could hold the water pressure necessary to run a collector system. The maximum allowable pressure inside the collector was not specified in the specifications section of our EDR. We chose a maximum pressure of 30 psi. This would be specified to the installer, and the system should never be charged higher than 30 psi. The water/antifreeze solution in the collector will be isolated from the domestic water side of the system, thus isolating the collector from higher well pump or city water supply pressures. In finding materials for our first preliminary prototype collector we ran into our first design dilemma. Our original design called for an aluminum collector. This was

due to the low cost of aluminum and the fact that our design eliminated the need for a highly conductive material. A full explanation of this statement is given in the EDR, but the graph below illustrates the point nicely. It shows that as the riser spacing approaches zero, the conductive efficiency of any metal approach 100%. Our design calls for a riser spacing of 2.25” on center, which results in a high conduction efficiency regardless of the material used.

Figure 2

Unfortunately, given our prototyping constraints, aluminum could not be used to construct the absorber plate. Aluminum pipes and fittings are not commonly used in industry and are therefore very expensive. The cost for these aluminum parts was prohibitive. Because of the prototyping nature of our project we decided to switch to copper piping. Our intention was still to use an aluminum sheet as the base of the absorber plate. After deciding to use copper pipes with an aluminum plate, we needed to devise a method of attaching the two materials. Because copper and aluminum are dissimilar metals, galvanic corrosion will occur over time at their junction. In order to prevent the two metals from coming into contact, we attempted to bond the pipes to the plate with a two-part epoxy designed for fixing leaks. The first prototype of this design is shown below.

Figure 3

After being pressurized, this prototype began to leak at around 15 psi. Because of the nature of galvanic corrosion and the inability of our first prototype to hold the specified pressure, our next attempt at prototyping our design involved a copper plate instead of aluminum. The exclusive use of copper also allowed us to change our bonding method. Instead of using expensive, insulating epoxy we opted to solder the half pipes to the plate. Our first attempt at soldering was unsuccessful. The pipe cutting process did not provide us with a clean enough edge to butt the pipe cleanly against the plate. The gaps that existed between the plate and the pipes were too big to fill with solder. Our third preliminary model was successful. In this design we stamped a channel into the foil plate before applying the half pipe riser. The diagram below shows a crosssection of the newly designed riser/plate interface.

Figure 4

Figure 5

This channel provides a place for the solder to flow and contact both the riser and the plate. This improved prototype design did not fail until water pressure reached 50 psi. Even at this pressure, the only leak occurred where we plugged the ends of the pipes with epoxy.

The table below summarizes the results of the preliminary experimentation and testing done early in the semester. Test #

Plate Material

Riser Material

Bonding Material

Failure Pressure

1

Aluminum

Copper

Epoxy

15 psi

2

Copper

Copper

Solder

----

3

Copper

Copper

Solder (in channel)

50 psi

Table 1

After successfully proving that our updated design could hold water at the prescribed pressure, we moved on to the construction phase. 2.2. Control System Design Initially, we had only the basic idea to use microprocessor and digital temperature sensors. We began our design process using a Motorola 68HC11 microprocessor MK11 lab kit with an LCD display and a BUFFALO Monitor operating system. We initially attempted to simulate the temperature sensor with a second microprocessor lab kit, the two kits communicating with each other. Our first inclination was to communicate using the Serial Communications Interface (SCI) of the microprocessor, as the digital sensors send and receive data serially. Our first prototyping was done with the SCI, attempting to get the first lab kit communicating with the second lab kit. This proved challenging, as the SCI system consists of a complicated network of registers, interrupts, and flags with various schemes for clearing the flags. Finally, with the two lab kits communicating effectively, we attempted to communicate with the actual temperature sensors using the SCI port. After spending much time without success, we determined from the datasheets and our experience that the sensors cannot function without a synchronized serial clock, which the SCI cannot provide. Therefore, we decided to utilize the microprocessor’s parallel input and output ports to create our own serial interface based on software. This allowed us to provide a synchronized clock to the temperature sensors and to read and write data at our own pace, a great luxury, we discovered. After toying with these ports and this communication system, we achieved initial communication with the temperature sensors. To our dismay, they proved faulty. Only one in four actually sent back a receivable signal, and this signal did not change with changing temperature. Thus, we

ordered new temperature sensors. With the new temperature sensors, we experience the same problem. Of the three new sensors that were soldered correctly, none returned any kind of data when called.

Figure 6 – MK11 Lab Kit

One possible explanation for the difficulty with the temperature sensors is that they turned out to be surface-mount and rather tiny. One chip was a mere 16 square millimeters in surface area. This gave us difficulties in manual soldering. Surface mount chips are intended for mass-production by machinery, not fumbly human hands. With a hot air soldering gun and soldering paste, we were finally able to solder the chips to circuit board adapters, then work with them on bread boards. After this process, we suspect that several of the chips had been overheated using this process, as they failed to function when tested. The second time around, we used slightly larger chips, which are somewhat easier to solder and less likely to overheat. Unfortunately, none of the larger chips functioned either. For our final design, we had intended to use a Motorola 68HC711E9 microprocessor version. The 711 indicates on-board EPROM, which would make the use of an external EPROM unnecessary, greatly simplifying the design. The only extra component necessary would be a 2 MHz crystal clock.

Figure 7 – Surface Mount LM75 Temperature Sensor

3. Implementation 3.1. Collector Construction The solar collector construction was completed in four major phases. Phase one was the cutting of copper tubes into the semi-circular shape prescribed by our design. Phase two was the manufacturing of our copper plate. Phase three was the bonding process. Both the semi-circular tubes and the manifolds needed to be connected to the collector plate. Finally, phase four included the absorber coating, frame insulation and installation of the absorber plate into the frame. We predicted that the absorber plate would be the most complicated construction aspect. This proved truer than we bargained for. Phase One: Cutting of Semicircular Tubing An important aspect to the semicircular riser design was the fact that the tubing is semicircular where it contacts the plate, but circular at the ends that stick out from the plate. This change in cross-section allows for an easy connection to the manifolds. Several manufacturing options were considered to complete this process, including the use of the band-saw and construction of a tool which, in similar fashion to a pipe cutting wheel, would latch on to the pipe and cut axially along the pipe. Finally, we decided to use the plasma cutter to slice the pipes. In order to guarantee straight cuts, we designed a jig to hold the pipe and guide the plasma cutter. This jig involved a cart that would hold the plasma torch and slide along a straight track at a fixed distance from the pipe. While this method demanded the construction of a complicated tool, we knew that an accurately dimensioned pipe was vital to the success of our project. The figure shows the manufacturing tool designed for the pipe cutting operation. Construction of the tool involved cutting the pieces with the band saw and machining them on the milling machine. The milling machine was necessary because exact dimensions were needed to guarantee precise cutting of the tubing.

Figure 5

The plasma cutter and jig did a fairly good job of cutting straight and accurately. However, the heat tended to warp the tubes slightly and it was necessary to thoroughly clean the pipes after they had been cut. After the longitudinal cut, the pipes were cut to length with a standard pipe cutter. Phase 2: Manufacturing of the Copper Plate We purchased a standard 3’x10’ 16-ounce copper plate from a local roofing supplier. The copper is normally cut for use as flashing for residential roofs. We had the copper cut to our specified dimensions by Obermann’s sheet metal shop because it was too large for the shear in our machine shop. Using a sizeable piece of scrap leftover after cutting our plate to size, we tested our stamping method. The new design called for a 1/8” deep channel to be stamped into the copper plate every 2-¼” on center (see Design Process section). This was done by using steel stock with a ¾” width. This size gave the 5/8” outer diameter pipe some space for the solder to flow. The pressing was done by placing the copper sheet over 2 lengths of steel stock divided by ¾” spacing and then stamping the ¾” stock over the top of the copper (see diagram below).

Figure 6

Phase 3 Soldering of Pipes and Manifold The soldering of the pipes and manifold was the most time consuming task in the whole project. Nearly three pounds of solder were used in the bonding of these pieces. First, all the pieces needed cleaning. A special wooden stand was made in order to hold the plate and act as an insulator so the heat from the torch would not spread quickly to the rest of the plate. The plate was clamped to the stand and then the pipes were clamped to the plate. A hand held torch was used to heat the plate and place the solder. The standard procedure for soldering the pipes included tacking the pipe down at several spread out locations and then filling in the gaps. Overall, the manifolds were easier to solder. The manifold was constructed using copper tee fittings and elbows where necessary. Because of the short distance between risers, the ends of the tee fittings were cut with the band saw so they could fit between the risers. Placing regular ¾” copper piping between the tee fittings, we then soldered the manifold onto the absorber plate. Gaps in each pipe existed where the semicircle tube changed to the full circular pipe. Rather than cut end caps for each of these small gaps as was planned in our original design (see EDR) we chose to use small rectangular pieces of copper bent at a right angle to cover the gaps. These rectangular angles were then soldered onto the bottom of the absorber plate and the exposed edge of the pipe. The figure below depicts this procedure.

Figure 7

Phase 4 Coating and Installation. After the absorber plate had been tested and found to be free of leaks, the SunEarth specialty coating was placed on the top surface. Painting was done outside and it was applied with conventional hand brushes. Approximately 0.5 liters were used to paint our collector plate. The installation of new insulation for our collector plate was a relatively easy section of the construction of the test collector. Before we could install the new insulation we needed to remove the existing insulation. This was done by using hand tools, including a utility knife and scraper. After removal of the existing insulation, a Celotex board of R-10 and thickness 1-3/8” was purchased. Several options were tested to determine the best way to cut the insulation into manageable sections for installation. The best method found was the use of a hand saw after marking the non-foil faced side with a scribe. Because the side insulation has a thickness of only ¾,” these sections were cut first into their normal dimensions and secondly into their cross-sectional dimensions. For installation into the frame the insulation was cut into sections and placed in the frame like pieces of a puzzle. This was done because the frame existed and the pieces had to fit into the dimensions of the frame. Because of the puzzle-piece nature of the installation, an additional expense of foil tape was added to our budget. The foil tape as used to seal the edging of the insulation. After the plate was coated, it was installed into the frame. New holes were drilled in the frame to allow the two manifold pipes to exit. Once the plate was installed, the glazing was cleaned and laid in place on top of the frame. The sheet metal angle that originally held the glazing in place was put back into position and fastened to the frame using self-tapping screws. Finally, for aesthetic purposes, the plate was painted bright blue. This extra step also easily distinguished it from the existing collector during testing at the Erikson Farm. 3.2. Control System Construction The construction of a working prototype control system could not be fully completed because of the difficulties with the use of the digital temperature sensors.

The construction plan for our control system design is still intact, though we were unable to implement it during the semester. For the control system hub, we would mount the 68HC711E9 microprocessor on a printed circuit board, using a breadboard adapter. Because of the on-board EPROM in the 711 version, the board would require no peripherals except for a crystal clock, which we would provide. The board would connect the relevant microprocessor output pins to wiring terminals, where temperature sensor wires could be attached. The entire board would measure approximately 3”x4”x1” and sit in a small plastic box with holes for leads. The temperature sensors would be mounted on the pipes, held down with a layer of thick insulation and taped in place using foil tape. Each sensor would be connected to the hub using four wires.

3.3. Operation After constructing the system, we installed the collector at the Erikson Farm. We set our collector up directly next to another collector of the same size. This existing collector was also a flat plate collector, but was constructed with full pipes spaced four inches on center as the collector plate. Setting these collectors next to each other allowed us to directly compare the performance of our collector to an existing model. Below is a picture of the two collectors set up side by side. Our collector is the one that is painted blue, on the left.

Figure 8

The collectors were oriented south, at an angle of approximately 45° from the horizontal. This angle is close to the latitude of the installation location. There were no obstructions to block sunlight from striking either of the collectors. Inside the turkey coop against which the collectors leaned, a hot water tank was installed to hold the water that would be heated by the collectors. This tank was a 40-gallon glass-lined storage tank. A laptop computer was also placed in the coop with National Instruments’ Field Point modules connected to record temperatures. The Field Point module uses RTD’s to sense up to eight different temperatures and then the computer can record these temperatures over time. Below is a photograph of the testing setup inside the turkey coop.

Figure 9

The temperatures that we monitored as we operated the system were the tank outlet temperature (bottom of the tank), the outlet temperatures of each collector, and the ambient air temperature. A circulator pump was installed at the bottom of the tank to pump water from the tank out to both of the collectors, which were piped in parallel. The water would flow through both, or either one of the collectors, depending on the test that we wanted to perform. The water would then return into the top of the tank. This simple test set up allowed us to perform a few different experiments. A schematic of the test installation piping is shown in the following figure.

Figure 10 During testing, the ball valves at the inlet to each collector were opened or closed to control the water flow through each collector. The circulator pump ran constantly during our testing. This was because our test installation did not have a means of removing heat from the system. As the water heated up during the day, the tank just continued to get hotter, since it is well insulated. Nobody was using the water in the system, and there was no radiator or other device to dissipate heat. So, by leaving the circulator on constantly, the heat that was collected during the day was radiated back out through the collector at night. This provided us with a way to remove heat from the system and prepare for another day of testing. Initially we closed the ball valve to the existing collector and set the flow rate through our collector to 0.4 gallons per minute. This forced all the water to flow through the new collector at a known flow rate. We set up the computer to log data and let the system run for several days. While we were setting up the computer, though, we had an interesting experience. The system was filled with water and the pump was running, but we unplugged the pump while we were setting up the computer. After a few minutes, the relief valve on the tank began to hiss a little bit and we heard a slight rumbling sound coming from the collector outside. The water in the collector was boiling! We quickly turned the pump back on and high-fived each other because we knew our system was finally making hot water after all those endless construction hours. A typical day of testing our collector is shown in appendix A. From this data we were able to determine that our collector had an efficiency of 57%. This is a good efficiency rate for flat plate solar hot water heaters. We calculated this efficiency based on the average insolation for an average April day in Harrisburg,

PA. This information was reported by the National Renewable Energy Laboratories (NREL), and is based on years of compiled data. NREL reports that for this area and this time of year, an average of 1250 BTU of energy is radiated from the sun per square foot each day. We determined from our data that our collector harnessed an average of 714 BTU per square foot in a day. This was figured from the fact that our system contained 20 gallons of water, and that water was raised 60°F over the course of a day. So, our collector utilized 57% of the available solar energy to heat the water. After a few days of recording data on the performance of our collector alone, we began to test the two units at the same time. We set the flow rates in each of the collectors to .4 gallons per minute by adjusting the ball valves. Since we knew that the flow rates were the same, we could monitor the outlet temperatures of each collector and get a direct comparison of output performance. Results of this test are also shown in the appendix A. Our collector was consistently a few degrees warmer than the existing collector. The collector design that we created actually improved the performance of the current technology.

4. Project Management Responsibility for the various aspects of the project was divided between the three members of the project team. Josh took charge of the collector, Jeremiah the controller, and Steve served as the project leader, coordinating the union of the two project divisions. Josh designed the main features of the collector worked on manufacturing the prototype. Jeremiah did most of the microprocessor programming and learned to work with the temperature sensor protocol. Steve focused primarily on the collector for much of the semester, using his manufacturing experience to do a lot of the manufacturing work. Toward the end of the semester, after the collector was nearly finished, he helped with the programming and construction of the controller. The Gantt chart on the following page served as our scheduling guide over the course of the project. It was very helpful in tracking the progress we were making and reminding us of the critical completion dates that we needed to hit. This updated Gantt chart shows both projected and actual completion dates for each piece of the project. As can be seen on the chart, the construction of both the collector plate and the controller took significantly longer than expected. After prototyping our original collector plate design, we discovered that our epoxy was not strong enough to seal the joints between the collector plate and the half-pipes. We then redesigned the collector plate, using solder to join the pipes to the plate. This process was much more labor intensive and time consuming than we had anticipated. Because of this extra labor and time spent on the collector plate, time had to be taken away from the controller. This was unfortunate, because the controller had construction problems as well. The temperature sensors didn’t work properly, and had a complicated and confusing protocol that took months to unravel. So, the construction of the collector and controller took far longer than anticipated, which pushed back our testing dates. In the end, we only tested for a week, rather than the month originally scheduled.

5. Budget The target cost for completion of our prototype was $300. This is the amount allotted for each Senior Project by the Engineering Department. While the final system design is intended to provide hot water for an actual home, the prototype is a smaller version, which cost less to construct. Some components were donated to our project, further reducing our costs. This small-scale prototype was intended to prove that our design would outperform a similar collector constructed with current technology. Below is a listing of the costs incurred during completion of the project. The total amount actually spent (not including gifts) was $324.18. This figure is very close to our target of $300. We were expecting to use an aluminum absorber plate, which would have been less expensive than the copper sheet that we needed to purchase. Also, the breadboard adapters that we needed to buy to work with the surface mount temperature sensors were an unforeseen expense. Overall, we were able to come rather close to meeting our budget. Project Expenses Item

Unit Cost

Quantity

Total Cost

Copper Absorber Plate

$1.83 / sq. ft

30 sq. ft

$54.90

Copper Piping

$.48 / linear ft

100 ft

$48.00

Solder

$7.76 / lb

3 lb.

$23.28

Misc. Pipe Fittings

$50.00

Collector Insulation

$.50 / sq. ft

32 sq. ft

$16.00

Specialty Coating

$25.00 / liter

1 liter

$25.00

Microprocessor

$16.00

2

$32.00

Temperature Sensors

$2

5

$10.00

Breadboard Adapters

$2

20

$40.00

Shipping

$25.00

Hot Water Tank

$250

1

$250.00*

Circulator

$80

1

$80.00*

Misc. Shop Expenses

Approx. $75

Total *These items were donated to the project free of charge.

$75.00*

$324.18

While our prototype system was constructed for around $300, the actual full-sized system design would be more expensive. Based on the pricing for the components used during our construction, and on estimates received from various vendors, we estimated the material costs of a full size 4’x 8’ collector to be approximately $1100. The budget estimate for our designed system is outlined in the following table. Budget Estimate for Actual Installed Unit Component

Price

Quantity

Total Price

Collector Box

$50.00

1

$50.00

Glazing

$7.00/ ft2

32 ft2

$224.00

Absorber Plate

$1.83/ft2

32 ft2

$59.00

Riser Piping

$0.48/ft

64ft

$31.00

Manifolds

$5.00

2

$10.00

Insulation

$0.50/ ft2

40 ft2

$20.00

Solder

$7.75/lb

5lb

$39.00

Microprocessor

$16.00

1

$16.00

Temperature Sensors

$2.00

4

$8.00

Misc. Electrical

$50.00

Piping

$0.48/ft

100ft

$48.00

Valves

$8.00

5

$40.00

Pumps

$80.00

2

$160.00

Heat Exchanger

$90.00

1

$90.00

Hot Water Tank

$250.00

1

$250.00

Antifreeze

$5.00/gal

2gal

$10.00

Total Cost

$1100.00

6. Conclusions The objectives for our project were as follows: 1. System freeze protected to –10°F and able to withstand temperatures up to 110°F, using only automatic controls (no manual maintenance). 2. Save greater than 40% of yearly hot water heating costs. 3. Comply with all government plumbing, building, and wiring codes. 4. Meet a $1000 cost target for installed unit. 5. Fully automate system to require no more than a yearly maintenance check. 6. Configure control system to allow for interface with different existing hot water systems. For the first objective, we came up with a design able to protect from –10°F to 110°F air temperature. The closed antifreeze loop would be the first step, using a propylene glycol antifreeze concentration that would protect to the low temperature without a problem. For the high temperature protection, the issue is that the collector fluid temperature can rise to near 300°F. At such high temperatures, over time, the antifreeze can begin to break down and lose its temperature properties, at which time it would have to be replaced. We remedied this with our controls system design. Whenever the temperature of drops below 0°F or rises above 180°F, the pumps will run constantly. This keeps the fluid exposed to the extreme temperatures moving, and prevents it from freezing or overheating. This provision will allow the antifreeze to last longer and allow us to work with a less concentrated mixture. The difficulties that we encountered with our control system prevented us from implementing and testing this in a real-world situation, but we wrote a microprocessor program to accomplish the objective and simulated the situation on the computer. It is a simple algorithm. We have no doubt that the microprocessor is able to accomplish this objective equally well in actual use. For our second objective, we have encouraging data for the time that we were able to test. In the week in April that we tested, we found that our prototype, if scaled up to full size, would have produced 116% of the daily hot water needs for a family of four. This is much higher, obviously, than the 40% goal in our objectives. One consideration is that the week that we tested for was sunnier than the average week in April. April is

also typicall an above average month in terms of available solar radiation. In addition, the heat produced over and above the family’s needs cannot be considered useful. We cannot average that into our calculation. Also, because we left the pumps run constantly, we cannot accurately predict what the implications of our controller would be, nor the impact of possible sporadic hot water usage. Because we could not test the system in its actual function, that is, mounted on a home and attached to a traditional water heating system throughout the year, we cannot guarantee that our system meets this objective. However, 116% for the week that we tested in April is extremely encouraging, and gives us great confidence that our system would go well beyond our goal of 40% yearly. The following are calculations for the percentage of home hot water needs for a family of four produced by a 4’ x 8’ collector. Hot Water Produced by collector system: 57% × 1250

BTU / ft 2 days BTU × 32 ft 2 × 30 = 684,000 day month month

Hot Water Needed for a Family of Four: 9

gallons / person days gallons × 30 × 4 people = 1080 day month month

Monthly Energy Needed to Heat Water: 1080

gallons BTU lb BTU ×1 × 8.33 × (120 − 55)° R = 584,800 month lb ⋅ ° R gallon month

Percentage of Monthly Hot Water Produced: 684,000 Btu Produced / 584,800 Btu Needed = 116% ~ 100% Hot Water needs for April

Our third objective is fairly straightforward. The system that we designed can be installed to meet plumbing, building, and wiring codes. It is made from components that, properly installed, meet government code standards. The heat exchanger may have to be varied from a single-walled to a double-walled version, depending on state regulations. Pumps must be selected for use with potable water. Otherwise, it uses only standard copper piping, standard wiring and junction boxes, and piping solder.

For mounting on the roof of a house, our design has the advantage of being a flatplate. Batch collectors can hold many gallons of water inside, posing structural problems. Flat-plates, however, weigh usually less than fifty pounds full, posing no roofmounting problems. We set our fourth objective, budget, for a prototype system of $300 and an installed system of $1000. The prototype rang in at $324.18, slightly higher than estimated. This includes all our material costs, except for gifts-in-kind. The reasons that we ran over budget were because of our design change from aluminum to copper and the need for temperature sensor adapters. The change to copper added $28 and the temperature sensor adapters added $20 to the budget. For the actual installed budget, the price tag is an estimated $1100. This reflects all the material costs associated with the entire setup. Most estimates are accurate within a few dollars. Where there could be some discrepancy between prices, we estimated high. The major factor that pushed our budget over our initial goal was the glazing. Low iron tempered glass turned out to be rather expensive. Even though we did not officially meet our budget objective, we still did well. In comparison with industry standards, $1100 is quite inexpensive. With labor and shop costs, of course, this price would rise considerably. With a little improvement in our manufacturing process, we expect that our system would be very competitive. Most systems currently on the market run well over $2000, not including installation costs. For a design with several unique features making it more efficient, $1100 is a great deal. Our fifth objective deals primarily with the control system. To fully automate the system means that no turning off or on of pumps will be required at any time throughout the year, no matter what temperatures arise. Our design meets this objective. The control system is designed to completely control the pumps, including over temperature and under temperature protection, so that there is never any need for draining in the winter or opening or closing valves in the summer. The built-in temperature protection also helps to extend the life of the antifreeze, so that it will have to be changed less often than a traditional system. There are no filters or moving parts within the system that will wear out or need replacing any sooner than in a traditional water heating system.

Unfortunately, we were unable to test the control system in its function, so we cannot determine whether it would be successful in this objective. However, the simulated program worked perfectly. With reliable temperature sensors, the control system would be able to fully automate the solar water heater. The sixth objective was a simple software consideration. The three primary methods for heating water are using a boiler, a heated tank, or an on-demand or tankless coil water heater. For a boiler and a heated tank, the algorithm is the same. Our microprocessor would take control of the traditional heater, turning it on and off based on the temperatures read from the sensor that we would place in the tank. On-demand and tankless coil heaters do not use tanks, so this function would be unnecessary. We would just not attach the sensor. The algorithm is written in such a way that, if the control for the traditional heater is not attached, it affects nothing else. We would simply leave control of the traditional heater alone. We could not test this in practice, because we were not able hook the system up to a traditional hot water heater and because our control system was not functioning in the end. However, based on the knowledge of one of our group members that worked previously as an intern in a boiler manufacturing plant, we know that the algorithm would work effectively with a traditional system. Through this project, we learned much. First, we familiarized ourselves with the operation and design of a solar water heater. Our design, in many ways, is on the cutting edge of solar water heating technology. No existing solar water heating systems use microprocessor control or integrated circuit temperature sensors. No solar systems currently on the market can take control of the traditional water heating system. We found no systems that make use of the direct heat transfer characteristics of metal like ours does. Through our project, we were able to work on the forefront of this burgeoning industry. As oil continues to become scarcer and dependence on alternative energy sources increases, the solar water heating industry will grow as well. Technologies like those that we worked with on our project will become more and more valuable as the years go by. Second, we were able to work through an entire engineering design project, much like one that goes on in industry. We learned that certain things take much longer than

you think they will. We had in mind constructing our collector in a couple weeks to a month, when in actuality it took three months. Manufacturing can be much more time consuming and difficult than it looks initially. We never imagined how difficult it could be to shear pipes in half, then solder eight-foot sections to a copper sheet. Heavens to Betsy! The controller provided us with many frustrations as well. Weeks of trying to program the microprocessor and constantly running into the same problems became difficult do cope with. In the end, it became apparent that we had to completely change the design. Reflecting on this frustrating situation, we learned that asking lots of questions of knowledgeable people, both on the front end and throughout the project is extremely useful. The major breakthroughs on the controller followed the various small pieces of insight that we received from fellow students, Earl, and Dr. Pratt. We underestimated the value of these small pieces of insight. Finally, we learned something about the dynamics of working with a design group for a long period of time. We each, at times, had different ideas about the best way to accomplish our goals and about which pieces of the project were worth devoting the most time to. This had to be negotiated and we had to learn to compromise. At the times when the project became most frustrating, we had to hold our tempers and continue to support each other even when we got on each others’ nerves. On the other hand, working with each other helped to keep us accountable and motivated. It made working much more fun at times. It would have been extremely difficult to work through the frustrating times alone.

7. Recommendations for Future Work While much was accomplished over the course of the year by our project team, some work remains that could be completed by future groups. Both the collector manufacturing process and the controller design could benefit from further developments. The design of the flat plate solar collector was shown to be an improvement over current technologies. The construction of this collector, however, is very labor intensive and requires much attention to detail. An easier, more efficient manufacturing process would need to be developed to make this product viable. While this project showed that the half-pipe design idea does work, a better method still needs to be developed to implement this design. We recommend a process that would involve using a second plate instead of individual pipes. This would eliminate the need to cut the pipes axially. The second plate could be stamped to include channels for water flow, and then bake soldered to the bottom plate. This process is illustrated in the diagram below.

Figure 11 Also, we used copper to prototype the system, but the original design called for aluminum. Aluminum is significantly less expensive than copper. Although it is less thermally conductive, aluminum could work well in our application because of the closer spacing of our riser tubes. Determining a method of creating the absorber plate with aluminum would decrease the cost of manufacturing. The future work for the control system basically deals with the temperature sensors. We had tremendous difficulty making use of the complicated protocols and small size of the digital integrated circuit sensors. The problems included soldering difficulty, which led us to use breadboard adapter chips in an attempt to make the process easier. Soldering to the adapters was nonetheless difficult and was one of the primary sources of difficulty. In addition, the adapters made the sensor assembly bulky and difficult to mount.

Because the 68HC11 microprocessor contains an on-board analog to digital converter, analog temperature sensors could be used to simplify the temperature data gathering process. The LM75 digital sensor’s analog cousin, the LM34 would take care of many of the difficulties we faced with the digital sensors. The analog sensors come in a package that would be much easier to work with, primarily because they require only three pins. The three long pins extend away from the package, allowing wires to be twisted or soldered directly onto them. The sensor could then be mounted inside the pipe, using a rubber tap with a tiny hole drilled in the end and filled with a rubber or epoxy to hold the sensor in place, keep water out, and insulate the sensor from the air. The sensor would send back a simple analog voltage, proportional to the water temperature inside the pipe. The analog to digital converter inside the microprocessor would convert this voltage to a number representing the temperature, which the control algorithm could use to control the pumps and relays.

1

Bibliography Energy Efficiency and Renewable Energy Network. Department of Energy. 9 September 2001. . “Federal Technology Alert.” Federal Energy Management Program. 9-10-01. Harrison, John and Tom Tiedeman. “Solar Water Heating Options in Florida.” Florida State Environment Commission. May 1997. 9-10-01. Hazeltine, Barrett, and Christopher Bull. Appropriate Technology: Tools, Choices, and Implications. San Diego: Academic Press, 1999. Lipschutz, Ronnie D., et al. The Energy Saver’s Handbook For Town and City People. Emmaus, PA: Rodale Press, 1982. “North Carolina Solar Center Factsheet. Passive and Active Solar: Domestic Hot Water Systems.” North Carolina Solar Center. Oct. 1998. 9-10-01. Olson, Ken. “Solar Hot Water: A Primer.” Home Power Magazine. 2001. 9-10-01. Solar Collectors, Energy Storage, and Materials. Ed. Francis de Winter. Cambridge, MA: The MIT Press, 1990. Solar Living Sourcebook. 10 ed. Eds. Doug Pratt and John Schaffer. Ukiah, CA: Real Goods Trading Co., 1999. Solar Radiation Data Manual for Flat-Plate and Concentrating Collectors. Solar Radiation Resources Information. 9 September 2001.