Originally posted November 15, 2011
Last Wednesday, November 9, I was forwarded a link to a blog post on www.oeic.us/ regarding a study released by Carl McDaniel and David Borton entitled “Evacuated Tube Solar Hot Water Systems Are Not Energy Efficient or Cost Effective for Domestic Hot Water Heating in the Northeastern U.S. Climate.” As a solar heating instructor at Kennebec Valley Community College in Farifield, ME and the owner of a solar heating design/build company, I was intrigued to review the study. I was dismayed to read a research paper that was fraught with practical and academic errors – one that took the liberty of extrapolating the data from one rather unique application in Ohio and applying it to an entire technology over a large region of the United States. Below is a summary of concerns related to this study.
A copy of this Response is available in .pdf format below in RESOURCES.
The system described in the study is owned by one of the authors, Carl McDaniel. The McDaniel house was designed to use minimal amounts of hot water. McDaniel had water meters installed on the inlet to his water heating appliance and found that the facility used 2933 gallons per year during a 27-month period between August 2009 and February 2011.
The solar heating system
The system utilized an Apricus AP-30 evacuated tube collector, an 80-gallon Bradford White electric water heater and a SolarHot SolVelox pump station. The solar pump station consists of two circulators – one for the pressurized glycol in the collector loop and one that circulated water from the tank through an external flat plate heat exchanger to transfer heat from the collectors to the storage tank. The controller for the system was a Steca TR301U differential controller.
A mixing valve was placed on the outlet of the solar storage tank to temper water before it prefed a Bosch electric tankless water heater. The intent of the design was to have the tankless water heater provide supplemental heat if the solar preheated water entering the tankless heater was less than 111F. The electric element in the Bradford White storage tank was disabled.
Figure 1 illustrates the average annual energy used to produce hot water as concluded by McDaniel and Borton from their data collection and analysis. Essentially, two-thirds of the contribution to the household’s water use was a result of the solar heating system. They estimated that the solar contribution was approximately 870,000 BTU per year of the 1,300,000 BTU required to heat water.
The research also found that the solar heating system required the equivalent of approximately 760,000 BTU/yr of pump energy to produce the usable 870,000 BTU from the solar preheat system
(see Figure 2). The authors argue that because of this inefficiency, the system was a poor choice for this residence and, furthermore, an inefficient system for use in the northeastern United States.
While the topic of McDaniel’s and Borton’s research – namely quantifying the overall effectiveness and economic impact of solar heating systems - is one of importance, their methodology and conclusions are fundamentally flawed. While it is clear from their research that the particular system installed on McDaniel’s residence performed inefficiently, the authors are in error in their attempt to make this case study indicative of evacuated tube technology across the northeastern United States.
Lack of representative sample
The McDaniel house is atypical. The average daily hot water use was determined to be roughly 8 gallons. The authors indicate that this limited water use is due to conservation measures in the home – low flow showers, a front-loading washing machine, and a dishwasher that uses 5 gallons per use. Based on the authors’ estimates of use per fixtures, the two residents of this house would be allowed to perform the following tasks weekly to equate to this water use:
- One load of laundry
- One load of dishes
- Four showers per person
This includes one extra gallon per week for miscellaneous hot water (handwashing, sink-washing larger kitchen items, etc.). The authors note that this is less than one-third of the use of a typical two-person household, citing a report based on an EPA study in Seattle. National projections for hot water consumption are closer to 40 gallons per day, which is five times the water use of the residence in the McDaniel/Borton study.
The average number of people per household in the northeastern United States is above 2.5. Thus, an average hot water use of 45 gallons per day is more indicative of the average population of the region. There are a number of conservation methods that could be implemented to reduce this usage, but these measures would have to reduce consumption by over 80% to equate to the household examined in this study.
I agree with the authors’ conclusion that the installed system was not the most cost-effective solution for water heating at the McDaniel house. With such low water use, other properly-designed options – such as a conventional water heater, a tankless water heater, or a heat pump water heater – would likely show better economic performance than solar water heating.
As the authors attempt to apply this lesson to residential installations in the northeastern U.S., they make critical errors. One, which is common to the layperson when comparing the solar potential of two separate sites in the country, is the conclusion that an evacuated tube system might be appropriate in Yuma, AZ due to the increased amount of sunshine in the southwestern United States. The amount of sun is a factor, but not necessarily the driving factor. If it were, Germany would be bereft of solar; instead it is a global leader. One primary reason is the cost of competing energy sources. Figure 3 shows the average residential costs of electricity in the United States in 2007. As is illustrated in this map, the average electricity costs in the northeast are substantially higher than those in Ohio – nearly double in some locales. When adjusted for the residential cost of electricity, the solar production per dollar in New York exceeds that in Arizona. Electricity is not the only heat source, either. There are major energy security issues – politically, economically, and ecologically - in states such as Maine, where oil is a dominant heating fuel.
This study assumes that the solar heating system and integration with the tankless water heater was designed and implemented correctly. There is no discussion about how the system could be made more efficient nor about concerns with the overall system design. The reality of this system is that it was poorly designed for the application.
The first glaring issue is the use of the Bosch RP17PT tankless heater. This model is flow dependent, meaning that it provides a fixed amount of energy (roughly 17kW) whenever the demand for hot water results in flow through the unit. It does not modulate the amount of energy used to heat the water – which is necessary for proper integration with a solar preheat system. The installation instructions for this unit explicitly state that water should not be preheated. There is a thermal cutoff switch that is supposed to shut off at least one element if the water entering the unit is greater than 85F. Since there is no discussion of errors with the unit or modification of the unit, this brings into question the reliability of the electricity usage results, which were indirectly used to determine the solar contribution. A Bosch technical representative verified that this use is inappropriate for the RP17PT unit. He was unable to determine how the unit would function in this situation given that at least one element should be disabled if the thermal switch was operational and the incoming water exceeded 85F.
The external pump station and controls were not optimally designed for minimizing parasitic loads. As mentioned previously, this design required two pumps. Based on the data, these pumps were drawing over 800Wh per day. With an average of approximately 5 hours per day for evacuated tube collectors, this equates to 160w for pumping. This is excessive. Controls and pumps that were readily available when this system was installed could have reduced the daily parasitic load to less than 150Wh per day without significantly increasing the installed cost of the system. If the desire was to eliminate these loads entirely, a PV-controlled system could have been used that would have been price-competitive and eliminated any parasitic loads.
Based on a conversation with the owner, the mixing valve between the solar storage tank and the tankless water heater did not have a temperature gauge installed. This is a critical feature in a solar heating system where additional heating may take place after the mixing valve. Without the ability to measure the outlet temperature of the mixing valve it is difficult to determine whether it is set appropriately. Certain models can be set between 70F and 145F. If the mixing valve is set too low – say at 90F – the solar storage tank could be well above the desired set temperature, but the system would still require supplemental heat to get it to the 111F design temperature.
The ratio of solar storage to collector area was extremely high for this application. This is a critical determination when attempting to gain a high solar fraction. If a residence is using 8 gallons of water per day, it is more desirable to heat a smaller volume of water than to store large volumes of water. Running the tank hot with evacuated tube collectors is not a huge concern from a collection efficiency standpoint, since the thermal losses in evacuated tubes allow them to perform reasonably well under such conditions.
- The sizing problem led to significantly more pump energy being used than required. As designed, the system would continue to run two pumps to collect heat when the tank had reached its maximum temperature. A zone valve was used to drain water from the storage tank when it got too hot, thus introducing cold water from the supply to cool the tank. There are a number of other options that could have been used to minimize this wasteful practice.
It is clear that the system installed at the McDaniel house was not efficient. It would be difficult to develop a solar water heating system that would provide the convenience of a tankless water heater in Ohio for $36.00 per year. I do not take issue with this conclusion.
The conclusion that “it is clear that an [evacuated tube collector] is not cost effective under present economic conditions for residential hot water in most situations and especially so in the northeastern U.S. climates” - is unjustified. To make such a deduction, the authors would need to use a representative case study that is properly designed and integrated. They did not. Other tests have shown remarkably different results.
More appropriate conclusions to this study would be:
- systems need to be designed and integrated properly to achieve maximum efficiency
- solar heating systems may not be the most cost-effective option in low-volume households
- parasitic loads must be considered when designing solar heating systems
Solar heating – like other modes of energy production – is not a one-size-fits-all solution. In fact, one obstacle in the technology’s widespread adoption is this characteristic. It is much easier to talk in specifics about a technology like photovoltaics or electricity because their contribution is more directly measurable. Solar heating, on the other hand, is impacted by the behavior of the individuals in the building, the efficiency of the auxiliary system, outdoor temperatures, and a host of other factors. During the past few years, a number of heat quantity measuring devices have become affordable, which will help create more accurate and accessible data for better measuring the efficacy of these systems. Empirical data from such sources is critical in determining reasonable conclusions to issues of efficiency and economics.
Vaughan Woodruff is a NABCEP-Certified Solar Thermal Installer™ and Solar Thermal Instructor at Kennebec Valley Community College in Fairfield, ME. He is a member of the NABCEP Solar Heating Entry Level Technical Committee, lead instructor for the Northeast Solar Heating & Cooling Instructor Training Program. He has a B.S. in Civil Engineering from the University of Maine and an M.A. in Education from Prescott College. He is currently finishing up the installation of his Econoburn gasification boiler that heats his retrofitted 100-year-old home, office, and workshop. He heats his water with the sun, as well.