One Community’s sustainable energy infrastructure is just one aspect of our blueprints for self-sufficient and self-propagating teacher/demonstration communities, villages, and cities strategy to be built around the world. Just as we will be showcasing a diversity of eco building methodologies and alternative food production options, so too will we showcase a diversity of alternative energy methodologies ranging from traditional generators to solar, wind, and newer technologies as they become available. Our initial system to supply for the energy needs of Pod 1, our aquaponics systems, and the Center of Peace community center is a tried-and-true 283 kWh photovoltaic solar power system.
This blog post is to open source the process used to design this system. For complete One Community energy details including pricing, assessment of other systems, and how we intend to meet our needs before this system is up and running:
CLICK HERE TO VISIT OUR COMPLETE SUSTAINABLE ENERGY INFRASTRUCTURE PAGE
Doug Pratt: Solar Systems Design Engineer
JP Novak: Power Backup Systems Designer at Native Construction and Renewables
Lorenzo Zjalarre: Physicist and Energy Efficiency Expert
Electric power requirements (see below) have been estimated by JP Novak of Build Native.com. The above system was then designed by Doug Pratt applying his 27 years of solar design and installation experience to confirm these estimates (below) seem reasonable. Estimating how people will use power in advance is always a guessing game. Humans are nothing if not variable and we anticipate that this system will almost certainly require some fine-tuning. We also anticipate that our group will learn from experience and probably become more energy aware and conservative with time. To help us gather data and fine-tune our process as part of our open source sharing, we will be using simple metering on all pods and the aquaponics systems. By doing this we will be able to identify “energy hogs” and share this data, our solutions, and the objective energy saving results of our solutions.
The total electrical use for Pod 1, aquaponics, and the Center of Peace community center, on a yearly average, is estimated to be 282.5 kWh per day. This is the figure used to size the solar electric system. In addition, all the AC appliances that were likely to be on simultaneously were totaled up. These included a percentage of lights, laptops, microwaves in the pods, along with all the aquaponics hardware, and most of the kitchen and community center lighting and hardware (including the hot tub). This max AC surge was about 76 kW and was the figure used to size the inverter pack.
Pod 1 Power Requirements (64 Units) | |||||
Appliance | Wattage | Hrs/day | Units | kWh/day | kWh/mo |
Light | 100 | 4 | 64 | 25.6 | 768 |
Laptop | 80 | 5 | 32 | 12.8 | 384 |
Hair Dryer | 1400 | 0.5 | 11 | 7.7 | 231 |
Microwave | 1200 | 0.5 | 11 | 6.6 | 198 |
Cellphone | 4 | 3 | 64 | 0.768 | 23.04 |
Total | 53.468 | 1604.04 | |||
Aquaponics Power Requirements | |||||
Pumps | 500 | 24 | 3 | 36 | 1080 |
Fans | 50 | 24 | 6 | 7.2 | 216 |
Air Pump | 100 | 24 | 2 | 4.8 | 144 |
Light | 100 | 2 | 4 | 0.8 | 24 |
Total | 48.8 | 1464 | |||
Center of Peace Power Requirements | |||||
(8 Suites/170 capacity dining room/3 Conference Areas/Laundry/Kitchen/Library) | |||||
Appliance | Wattage | Hrs/day | Units | kWh/day | kWh/mo |
Satellite Dish | 50 | 24 | 2 | 2.4 | 72 |
Computer | 300 | 2 | 15 | 9 | 270 |
Multi-media Other | 250 | 2 | 3 | 1.5 | 45 |
DVD Player | 50 | 2 | 3 | 0.3 | 9 |
Stereo/Music | 1000 | 2 | 3 | 6 | 180 |
Lighting | 100 | 4 | 50 | 20 | 600 |
Hot Tub | 20000 | 4 | 1 | 80 | 2400 |
Maytag Washer (Maxima 4.3 cuft) | 200 | 2.5 | 5 | 2.5 | 75 |
Maytag Dryer (Maxima 7.4 cuft) | 1000 | 2.5 | 5 | 12.5 | 375 |
Refrigerator (40 cuft) | 1000 | 6 | 2 | 12 | 360 |
Vacuum Cleaner | 1000 | 0.5 | 2 | 1 | 30 |
Walk-in Freezer (8′ x 6′ x 8′) | 0 | 2280 | |||
Walk-in Cooler (10′ x 20′ x 8′) | 0 | 1600 | |||
Dishwasher | 2000 | 4 | 1 | 8 | 240 |
Stand Mixer (30 qt.) | 2000 | 1 | 1 | 2 | 60 |
Griddle (3′ x 2′) | 1500 | 2 | 1 | 3 | 90 |
Oven | 10000 | 2 | 1 | 20 | 600 |
Total | 180.2 | 9286 | |||
Grand Total | 282.468 | 12354 |
Okay, so how do we go from kWh per day to PV arrays on the ground, and battery sizing, and inverter sizing, etc.? Here’s how Doug described it for us:
Looking at available sunlight for the property: The National Renewable Energy Labs (NREL) went out and measured sunlight availability for several hundred sites across the U.S., and they did it for 30 years. 1960 thru 1990. So we’ve got a nice average. This is the “NREL Redbook”, and is the standard source for estimating sunlight availability at any point in the U.S., for any time of year. For our location we have a yearly average of 5.9 hrs of peak sun per day. Peak sun? What’s that? That’s the scientific definition of full sunlight on the Earth’s surface. A “full sun” is defined as 1,000 watts per square meter. Now it’s immediately apparent that’s an impossibly round figure. And you’re right. Reality on the ground varies widely depending on humidity, altitude, sun angle, and a host of other variables called “the weather.” What NREL has done for us is to take all the hours of sunlight on a particular site and condense it down, as if all the hours were at perfect solar noon – 5.9 peak hours in this case. Which is pretty handy, because PV modules are rated to produce a certain amount of power at “full sun.” If we know a site averages a certain number of peak hours of sunlight, we can closely estimate how much power a given PV array will deliver. Now, a warning here, we’re talking about the weather. And it varies from year to year. In fact the NREL data clearly demonstrates that it varies by plus or minus 9% yearly. So it’s not worth getting too gnat’s ass accurate with our system sizing, as there’s bound to be yearly variations.
5.9 hours is the yearly average sun for our location. In December, at the lowest, it drops to 4.4 hours, which is still pretty good as solar sites go. For a December site, it’s excellent, and we’re going to use the 4.4 hour figure for PV system sizing. Now we know how many kWh per day your complex needs, we know what the average sun is going to be in December, what’s left is system efficiency. How much is lost to wiring, dusty modules, batteries, inverters, etc? Real world measured efficiency for battery-based systems ranges from 50% to 70%. Since much of the energy in this system will be used during daylight hours and will not need to be stored in batteries, I’m giving this system a fairly high 65% efficiency rating. This is completely seat of the pants estimation based on experience with large battery- based systems.
So we’ve got a 282.5 kWh nut to crack with 4.4 hours of peak sun and a 65% efficient collection and delivery system.
282.5 kWh / 4.4hrs / 65% = 98.77 kW of PV required. How many of what PV module is left until later, probably until right before purchase as prices and module brands have been shifting rapidly.
Batteries are the largest expense for the system. Lead prices just keep rising as the world becomes more industrialized. Lead-acid batteries still represent the best buy for remote systems. (And before you ask, lithium-ion batteries are still at least 4-5 times more expensive, and haven’t proved they’ll last longer than lead-acid. Who hasn’t had problems with phone or laptop batteries?)
When sizing off-grid battery packs we usually aim for about 2 to 3 days worth of storage capacity. Less capacity means the batteries get cycled deeply on a day to day basis, which isn’t good for life expectancy. More capacity raises the cost to where it’s cheaper to start the backup generator to meet the occasional shortfall.
Batteries are sized by amp-hours rather than watt-hours, so we have to divide our watt-hour figure by the battery voltage – 48-volt in this case. (If you remember your high school physics, watts divided by volts equal amps. Or volts times amps equals watts.) We also have to factor in how deeply we’re willing to cycle our batteries. The true deep-cycle batteries we’ll be using will tolerate cycles down to 80% depth of discharge (DOD), but again, deep cycles aren’t good for life expectancy, so we’re going to draw the line at 70% DOD. Considering the high quality of the Unigy II batteries we’ll be using, along with reasonable cycle depth, this battery pack should enjoy a 15 to 20 year life expectancy. By which point lithium-ion batteries may be a better choice. That’s a bridge to cross when we get there.
282,500 watt-hours x 2.5 days / 48 volt / 70% = 21,019 amp-hours @ 48v. This is one honkin’ BIG battery! To help make it more manageable, we’re going to use an SMA Sunny Island Multi-Cluster inverter package which divides the inverters up into four separate nodes, with each node having its own battery pack. And that brings us to…
Doug chose the Sunny Island Multi-Cluster inverter package for several reasons. It’s highly reliable and adaptable German engineering at its best. It consists of 12 individual Sunny Island 6kW inverters wired as four groups of 3 inverters each. So 12 x 6kW = 72kW, very close to the max AC surge requirement we estimated earlier. (Each 6kW Sunny Island can deliver 8.4kW for 1 minute, or 11.0kW for 3 seconds for true surges.) Each node of 3 inverters will cover the A, B, and C phases of your 208vac 3-phase system. As power demand increases, the Multi-Cluster will activate more nodes as needed. So we won’t have a lot of inverter capacity turned on, using power, and just waiting for something to happen. Capacity will only get turned on as needed. Each node has its own battery pack, which will make the individual packs more manageable. And if any one inverter or battery pack needs service, that node can be shut down, while the rest of the system will still operate normally. Also, the Sunny Island system uses conventional high-voltage grid-tie inverters to process the incoming PV power. So transmission from PV arrays hundreds of feet away are much less of a problem.
While this system is designed to be largely automatic and self-sustaining, there will be one or more designated maintenance and service personnel for the community. This person will be in charge of system operations, and trained to be familiar and very comfortable with the Sunny Island system. In addition, a great deal of system automation is possible with the Sunny Islands. As battery state of charge drops to critical levels, the Island can initiate start-up of backup generators, and/or shut down selected loads (the hot tub for instance). Routine maintenance includes cleaning PV arrays, snugging up battery cables, and monitoring the Multi-Cluster for any warnings or problems. For this reason, someone dependable and knowledgeable will be “in charge” of the system at all times.