Solar Absorption Chillers
Architectural Engineering Masters student here (focusing on the mechanical/electrical side of buildings) and we are working on a small, very low carbon, energy efficient home that will both be built and eventually lived in but also is meant to showcase new ideas and apart of a competition.
Has anyone seen or had any experience with Solar Absorption Systems here in the United States? I have been fascinated by Purix (https://www.purix.com/experience/) and their Absorption Chiller and using a Radiant System for heating with a “side-arm” to help drive the regeneration when the sun isn’t shining.
We are likely to be off-grid so could probably get away with the 50Hz electrical requirements on the motor, but would love to know if there was any products here in the US available? Maybe it’s the wrong idea but have been very interested in the larger silica based aDsorption chillers for commercial systems and was interested in small scale ones. I know a lot of RV’s use absorption chillers but anyone have any experience? Additionally I know this would be adding complication and there’s the whole issue with maintenance but we are a bunch of engineering students ambitious and interested.
Ideally we would love to use something like the Chilltrix…..although our heating/cooling loads are only about a ton so they are much to large for us.
The whole goal is to get away from high GWP refrigerants for cooling so if anyone has any thoughts or ideas would love to hear them.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Absorption chillers are usually used in one of two places: either where there is a source of "waste heat" from something, like an engine or industrial process, or where there is a source of very cheap energy in the form of heat (such as in a city with a muncipal steam system). Absorption chillers tend to not be particularly efficient in other areas. It IS possible to run them off of an energy source like natural gas though, but they would only make financial sense if the natural gas was cheap enough to make the overall process cheaper to run than an electrical chiller.
I've only ever design with lithium bromide chillers, but the basic design considerations are going to be the same for other types. I've looked into using these absorption chillers for natural gas fired cooling systems, but the numbers have never worked out in terms of operating cost. The one place I HAVE found these chillers to make financial sense was in a combined cycle plant (I have been involved with the design of two such plants), where the absorption chillers allow for the ability to do building cooling using heat scavenged from the engine's cooling and exhaust systems. This heat would otherwise be wasted in a radiator to cool the engine, so the absorption chiller allows us to do something useful with that otherwise wasted heat, which ultimately brings the overall efficiency of the plant up. Unfortunately, I have had very few design oppurtunities like this.
Any solar-powered system is going to have similar system effiency concerns compared with a conventional electrical cooling system running from electricity produced by solar voltaics. If the overall system efficiency with the absorption chiller is lower than a comparable sized electrical system, you'd be better off going with the electrical system. Remember too that the electrical system will always be both more flexible, since electricity can be used for all kinds of things, and easier to maintain since you'll have more people in the field familar with the operation of the electrical system.
Do not neglect operations and maintenance costs for these types of systems. If the ongoing costs to keep a system running are too high, you'd have been better off with a different system in the beginning. This isn't only about the financial bottom line, either -- excessive ongoing costs are also implying that the money spent on the problematic system would have been better spent on other systems that may have contributed more to the overall efficiency of your project as a whole. Try to make sure that your projects overall energy use is minimized, don't focus too much on just the refrigerants used in the cooling system.
Bill
Hey Bill,
Thanks so much! Really insightful helpful comments and I especially appreciate the note on efficiency. I had been sort of so focused on how interesting and cool the product was I didn't think about the efficiency when the "waste heat" from solar thermal wasn't going to be around. Additionally why absorption chillers are typically always used in Co-Gen systems or RV's where the have the waste heat from the engine and just overlooked efficiency.
We are still in the early stages and working on the energy model but you brought up a lot of good points to consider. To give a little more background: Due to the size of the place, insulation and "hopeful" ACH, and just some other shading/strategies our energy use is very low especially during the daytime. Our DHW load outweighs the heating and cooling and so the thought was to use a small solar thermal system to supply DHW and heating during the day with a backup heat pump hot water heater. We are in a mostly 68% HDD dominant climate however want to plan for rising CDD's. We do get a lot of solar radiation though and especially in the summer and CDD season.
Additionally we are looking to build not just a NZE home but a Zero Carbon or very close to it home. We are doing a thorough Life Cycle Analysis and I am also pretty passionate about the problem with refrigerants. That being said I appreciate the comments and points about maintenance and ongoing costs and overall energy use versus just trying to get rid of refrigerants. Maybe solving one problem but not fixing the others so I appreciate the point.
We are working to be 100% Off-Grid at the moment. Additionally using a Hydrogen Storage System versus batteries for long term seasonal storage. Ideally the initial thought and goal would be to figure out how to engineer and use the waste heat from producing hydrogen and running it through the fuel cell to drive the absorption chilling process. Maybe over ambitious but we are engineering students so have to start high.
A large portion of our goal is to showcase "feasibility" and what can be done. We still want to be practical, realistic, and repeatable—but we also want to break the box a bit. That being said we don't want to be "different" without any real energy efficient, economical, or carbon benefit.
If you use solar voltaics, it's easy to shift power to cooling instead of producing DHW (or vice versa). There has been discussion on GBA in the past about solar collectors for DHW, and the consensus has usually been that solar voltaics and a heat pump hot water heater are a better way to go, or electric resistance if you need somewhere to "dump" excess solar production during peak production times of the day.
You could use an absorption chiller as a "dump" load too, with solar collectors, but you could do the same with solar voltaics and a relay or smart thermostat that would overcool a home if excess solar was available, and then "coast", using the home itself as a sort of thermal battery. Using your existing thermal mass as a thermal battery like this is extra green in a way, since you're not using any new materials or systems -- you're only running what you already have in an intelligent way to get some energy storage in a creative fashion. I've done this myself for years to take advantage of "time of day" electric rates: I intentionally overcool my home in the hour or so before the more expensive on-peak electric rate beings (at 11am for me). This means it takes longer before the temperature of the house rises enough to need the A/C to run again, so I end up running the A/C less during the on-peak hours. The great thing about this is that you don't need any batteries or other energy storage, the home itself IS your energy storage. All you do is change your operating scheduled and thermostat setpoints to make it work. You can do the same thing with heating if you're trying to maximize your use of solar production.
I think you'll find that hydrogen storage is incredibly inefficient. It's also more dangerous than batteries. I don't see this as a practical means to store energy in this way. I'd avoid moving forward this part of your system plan. If you want energy storage, and you will if you're off grid, batteries are a better way to go. I would put the batteries in a dedicated room, built as a fire rated room, and ideally I'd put a hydrogen sensor and exhaust fan in there as a safety precaution. This is how I design battery rooms commercially, although at a much larger scale (some battery rooms I've designed are several thousand square feet).
If you don't HAVE to be off grid, I'd go with an on grid, grid-tied system. This avoids the need for any long-term onsite energy storage (batteries, etc.). I often advocate on here to use solar for "peak shave" instead of trying for "net zero", since with a peak shave system, you are only offsetting your own consumption during the time your solar panels are producing -- you use grid power the rest of the time. This allows for smaller, cheaper, systems, so more accessible for many buyers. Peak shave systems are also practical regardless of what the net metering rules may be in your area, or even if there are any net metering provisions in your area at all.
You may have different goals though. I am usually looking for cost effective solutions that are within most people's budgets, which encourages large scale deployment. If, for example, 1,000 people were to install 3kw peak shave solar systems, you're going to remove a lot more total load from the grid than if you had a far smaller number of larger "net-zero" systems installed. I am always saying to look at the big picture, the "whole system" view, which sometimes results in different solutions to achieve a goal. Bringing down the cost of new systems helps to get them out there, which is better than trying to push very complex, expensive, systems that will only be used on a small scale. I also tend to focus more on energy efficiency than eliminating the use of any particular product or material. Different design goals, but similar visions for the future, I suppose :-)
Note that absorption chillers usually need pretty high temperature heat sources to be able to operate efficiently. Steam systems by definition will be over 212F, for example. Engine water jackets in cogen plants are usually around 180F or so, and the exhaust manifolds would run at over 900F if we'd let them (we don't -- we use the heat exchangers to scavenge this heat to run the absorption chillers). I don't think a hydrogen system is going to get you high enough temperatures on your cooling loop to be able to run an absorption chiller effectively.
BTW, in my datacenter designs at work, I have MASSIVE amounts of low-grade waste heat. I have facilities that have over a megawatt of waste heat ALL the time (24x7), but it's low grade, maybe 80-85F at most, since it's coming from all the servers in the data rooms. I am always looking for ways to do something useful with that heat. Usually we run cooling systems year round. I joke that it would be ideal to have a buisness next to us that did something like commercial towel drying or something that could use huge volumes of warm, dry air. For building heating, the 80ish degree air isn't hot enough to work efficiently. The slightly warm water from the hot side of the chillers isn't hot enough to be useful either. The best I've been able to come up with is to put coils of PEX under the parking lots, then use the waste heat to warm the parking lots in the winter to eliminate the need for plowing or salt. No salt means no salty runoff, so that's an enviornmental plus, and no plowing is a cost savings to the facility operator. The cooling loops improve system efficiency for the chillers, which is another plus in terms of operation efficiency and energy savings.
I mention the parking lots to encourage you to encourge your engineering students to think outside the box. A lot of breakthroughs have started in unusual ways, or in things that didn't initially look promising. My two favorite examples are the first transistor at bell labs, which looked likea high school science project put proved the theory was sounded and justifier further research. The second is the guy primarily responsible for inventing the LCD -- the liquid crystal display that we now use in phones, monitors, all kinds of things -- was almost fired for "wasting his time on a useless technology". I don't want to discourage you from following unusual lines of thought with systems designs, just be sure to keep in mind the practicality of those systems for large scale deployment, and ease of operations for end users, which are the two big things that help with adoption and make a big impact.
Bill
I work with a bank that finances data centers and I have been asking for a while why the industry has not more fully adopted all sorts of practices to make them more energy efficient: tighter building envelopes, large solar installs, and better use of waste heat, as you suggest. I suspect economically the dollars involved are not enough to spur more interest by senior management/private equity owners in these practices. What do you think?
They have adopted many things to make the facilities more efficient, but the requirements are different than in homes, so some things don't always make sense to residential builders.
The building envelopes are usually pretty tight, but the primary concern is usually moisture and dust. It's not uncommon to run the buildings (or at least the data roms within the buildings) at a slight positive pressure so that all the air leaks leak OUT instead of IN so that the enviornment can be controlled within the rooms. There are commercial makeup air requirements that sometimes have to be met too, or exhaust air requirements (such as a code of 1CFM per sqft of battery room space, which I fight with periodically so that we don't waste conditioned air -- it's more efficient to use hydrogen sensors and interlocked exhaust fans so that the exhaust air is only run when a hazard exists, which is how it's done in places like ice rinks that use ammonia-based refrigeration systems).
A LOT of effort goes into optimizing airflow and partitioning the room into hot and cold isles, to prevent air mixing. I have seen efficiency gains from doing this that are voer $1,000 per month in mid- to large size facilities.
Solar doesn't make any sense for a typical datacenter facility. There is not enough space to install enough solar collection to make any appreciable impact on a typical facility's electrical load, so the systems become an additional O and M expense that can't be justified. There are more efficiencies, and corresponding energy savings, to be gained optimizing the cooling system instead, which tends to be where most of the effort goes after the facilities are built.
The management of these facilities, and the owners are ABSOLUTELY FAR AND AWAY more concerned with energy efficiency than even most of the builders on GBA! The reason is that energy costs are the dominant operational cost for these facilities and no other expenses even come close. I have customers with electrical bills well over $100,000 PER MONTH. If I can save them 1% on those bills, they are saving $12,000 per year. EVERYTHING focuses on minimizing those energy costs, which means reducing consumption as much as possible. I have gone so far as to change electrical equipment layouts to shorten wire runs (which reduces resitive losses, and savings there also reduces cooling load, which is more energy reduction). We have painted roofs white (which makes a measureable improvement). We have had custom pitch fan blades manufactured for us to optimize the fan's load to the motor driving it, to reduce energy costs in that motor.
To give you an idea of how seriously this stuff is taken, it's not uncommon for me to do contract efficiency improvement projects where I bill no hourly rate. I make my money off of a percentage of the savings the project gains the site owners over a period of time. Usually that means I get some percentage of the annual energy savings dollars, paid to me quarterly, for 3 to 5 years after the project is completed. Typical projects save the operators $15-20k per year or more.
Can you think of any other builders that optimize the paint colors they use for energy efficiency? :-) Since energy costs dominate the budgets of these facilities so completely, the management and owners are very, very interested in doing anything they possibly can to save energy. Those cost savings are a huge motivator to constantly try to improve system efficiency of the facilities.
An add too: there are consulting engineers like me who specialize in these places. We tend to travel a lot (I do a lot of international work), and we seek out facilities that can be improved, and draw up presnetations to describe what we can do to help them be more efficient. We're like energy auditors, but we often do the audits for free, since we make our money either implementing the systems or from a percentage of the savings our systems provide. This means that even if the facility owners didn't think of something, there are consultants who make bring potential efficiency savings to them unsolicited, in the hopes of being contracted to put the new systems in place.
Bill
That all makes sense. I have seen the load calculations in terms of electricity bills as that plays into the economics of the transaction, although typically in the projects I see these costs are passed through to the lessees. I would have thought a geothermal cooling system would be a good investment, particularly if the excavation part was done at the same time as the foundation excavation. Have you designed any W2W systems as part of a data center project?
I don't see how you could make a geothermal system work at this scale, with the land area available. The problem is datacenters are VERY high power density compared to normal structures of comparable size, so you wouldn't have enough land area to put in enough wells to support a suitably sized ground source heat pump to support the thermal load of the building. A ground loop would be even lower performance than the wells.
We DO use water for cooling systems though. I usually design the larger facilities to use evaporative cooling towers, which are like a big box with a fan on top and a waterfall inside. The evaporation helps with cooling performance, and allows the return water to actually be below ambient temperature under certain conditions. These cooling towers work very well in dry areas (higher humidity levels reduce their efficiency), but they consume water at a pretty high rate so they are more expensive to run in areas of the country where water is expensive. If we can't use an evaporative cooling tower, we use what is known as a "drycooler", which is basically a very large radiator similar to what is in your car. Sometimes we'll just use DX cooling, which means the refrigerant loop condensor coil is what is located outside. There are pros and cons to all of them, so we pick whichever is the best fit for the facility we're designing, and the climate around the facility plays a big part in that.
I did consult on a project some years back that used their ornamental pond and fountain as their "cooling tower". They remarked to me that that's why they ran the fountain all year long, and had no freezeup problems in the winter -- the fountain was from the chiller, so the water was somewhat warm. I thought that was pretty clever. Very large facilities can use nearby rivers or lakes similar to how power plants sometimes handle their cooling, but I'm not aware of any facilities like that myself, and they require some special permits to use natural bodies of water for cooling.
If a lessee is leasing a facility they didn't build, and they didn't work out the operating costs, then they aren't a smart operator. I have been called in to consult on lease negotiations for facilities of this type many times, and part of my job in those cases is to audit the equipment and make recommendations as to operations and maintenance costs. Tenants can negotiate lower lease costs to offset less efficient facilities, which also tend to be older facilities. Sometimes the cost savings using older equipment that is preexisting is enough of an offset to justify higher ongoing operations costs, which in a way is a sort of recycling -- you're not installing new equipment. The tradeoff is that at some point in the future, your O and M costs can exceed your lease rate savings, and then you're in the red again. Part of my job in these situations is to advise on what I call a "cost horizon", which is a projection into the future of when those O and M costs will do what, and specifically when they will eclipse any savings from using preexisting equipment. I allow for equipment failures, ongoing maintenance, and other things in these assessements.
Most of my projects are building facilities for customers that will use the entire facility (private datacenters), or building facilities that will be operated to lease out to small customers on a per-rack or per-cage basis (public datacenters). In either case, the facility owner has an incentive to keep O and M costs down. With private facilities, you minimize the operating expenses for the buisness, since the datacenter is just a "cost" to them -- their primary buisness is something else (a large mortgage company is a customer of mine, as an example). Datacenters leasing services to the public can improve their profit margins by improving efficiency. They may build more cheaply in the beginning, but down the road efficiency upgrades let them increase their profits (or lower prices to attract more customers), so they still have a big incentive to be as efficient as possible.
My pricing model I sometimes use for my services also allows for no upfront costs to the systems operators, which is way for me to encourage them to try my services -- they have no downside, since I only charge a percentage of the savings they get from whatever efficiency upgrades I put into their facility. This encourages them to be more efficient, since they can only save money with this pricing model, they can never lose.
Bill
What’s the estimated cooling load? Could your well be pumped through radiant tubing in a slab to provide cooling? If you are going to have hydrogen generation via electrolysis you will probably have significant water needs and some storage/treatment. The heat pump water heater will also provide cooling and dehumidification.
The estimated cooling load at the moment is 8200 Btu/Hr or 2.4 kW. The original design was using one ductless mini split since the heating load was only about 7500 Btu/Hr or 2.2 kW but one of the challenges we are trying to avoid is the higher peak load from a heat pump for heating and cooling and the heat pump for DHW at the same time.
Fortunately the water consumption is actually pretty minimal ~400 mL/Hr.
Additionally we are using a split heat pump hot water heater. I had been curious about possibly having a ducted bypass to provide cooling when we needed it and to bypass it out when we didn't. I thought though however it would be dependent on hot water usage so it would be like "hey go take a shower its hot in here" and wasn't sure if that would be too advantageous.
Deleted
What climate is this for? If you are building a well insulated house in a humid climate, your sensible cooling load may be small compared to your latent cooling load. In that case, you'd have to make sure that if you were going to cool with chilled water from some alternative chiller type, you got low enough temperature water to provide good dehumidification. Or, you might want to think in terms of options for dehumidication primarily, with cooling as a secondary consideration. Solar desiccant dehumidification is a good one to consider, either liquid or solid desiccant. If you have a good dehmidification system, you can then use a small PV powered high-COP compression chiller, operating with a higher-than-normal evaporator temperature for good COP, since you don't need dehumidificaiton. Or you can use direct or indirect evaporative cooling.
Also take a look at the recently completed Global Cooling Prize competion for some innovative ways to combine some of these types of approaches.
If you want a chiller with a low GWP refrigerant, lots of refrigerators are sold with R600a now. You might be able to buy one and add your own heat exchangers to get the heat and cooling where you want them.
While this is not exactly what you are looking for, have you explored the solar assisted heat pumps by SAHP-a UK company? I know the Radiant Store in Troy NY carries them, but I do not have personal experience with them. It seems like these are interesting, relatively simple hot water systems that have been deployed in the UK and here with some success