We’ve been dedicating several posts to building an off-grid energy system with solar panels. Any off-grid energy system, especially solar, is best operated with batteries to store energy, unless you plan on making a generator available 24/7 to fill in the gaps. There are many available battery technologies, and eventually we’ll review many of these. For this post, we are going to focus on lead acid batteries, which are the standard by which all others are compared, good or bad. In later posts we’ll cover nickel-iron (or Edison) batteries, lithium iron phosphate, and other options.
Lead Acid Battery Types
The most important distinction when selecting lead acid batteries is whether it is a starting battery or a deep cycle battery. Deep cycle batteries can also be known as marine or golf cart batteries, but all of these are effectively the same for our purposes here. A starting battery is designed to supply a large current for a short period of time, a feature known as cranking amps. Conversely, a deep cycle battery is designed to supply a more moderate current for a long period of time. The physical construction of the plates determines which kind of battery it will be.
You should design your off-grid system using deep cycle batteries, so we’ll discuss this type in detail below. In a pinch, however, you could apply a large number of starting batteries, but you will have to be careful to not drain them any more than a typical heavy cranking cycle. Managing starting batteries in this way is beyond the scope of this article.
Deep cycle batteries are typically available in 6 volt, 8 volt or 12 volt options. Because of how batteries are made, the lower voltage batteries will generally contain more energy storage per cubic foot, and thus storage per invested materials. As a result, the lower voltage batteries, all other things being equal, will usually cost less per watt-hour. See the table below for some examples, prices are taken from Sam’s Club’s website:
|Model||V||20-A-hr||Wh||Price||Price per Wh||Price w/ $18 core||Price per Wh|
Note that the general rule of less-expensive per watt hour breaks down with the 8 volt GC8 when compared to the more capable ECG2. We’ll refer back to this table in later sections.
As you will recall from previous articles, we arranged solar panels in strings to reduce power losses by keeping currents low and voltages high. Similarly, arranging batteries in strings will accomplish the same goal. Many chargers output a maximum amount of current, more or less independent of the battery voltage. Simply by increasing the battery array voltage by connecting batteries in series, the charger becomes that much more effective. As an example, a 20 amp charger can store 240 watts in a 12 volt battery, but that same charger can store 480 watts in a 24 volt battery array. Not all chargers can support all voltage options, but most high quality chargers will support two or more.
For starter purposes, we recommend 24 volts for a battery array. Most chargers support both 12 volts and 24 volts anyway. Inverters are usually designed around a specific voltages, and are priced similarly for 12 volt and 24 volt options. At higher voltages, the prices start rising rapidly; a typical 48 volt inverter is generally much more expensive than a similar 24 volt version. If you know you are going to build a very large system, then choose 48 volts from the start.
We like using 6 volt and 12 volt batteries, and avoid 8 volt batteries, even though these can also be used to build a 24 volt array. 12 volts is such a common and useful voltage that having the option to string two 6 volts together is a nice bonus. One example could be repurposing 6 volt batteries nearing the end of their service life as gigantic uninterruptible power supply (UPS) batteries. Those UPS inverters, while square wave and of no use driving inductive loads such as freezers, are very handy for other loads such as computers and many 110 volt LED lights. Plus, the native 12 volt option itself is well supported by a variety of existing RV accessories. 8 volts winds up being a strange niche.
Cranking amps and cold cranking amps are common battery specifications, but are practically meaningless in the context of an off-grid battery array. If your batteries ever experience any usage close to the cranking amps rating, something has gone terribly wrong.
Lead acid batteries absorb about 70% of their charge over a period of about 2-3 hours, with the remaining capacity charged in another 8-10 hours. The first stage of this process is known as bulk charging, while the second stage is known as float charging. In some contexts, the term absorb will be used for this stage, with float reserved for maintaining a fully-charged battery (we prefer the three-stage terms bulk, absorb and float). In an off-grid system, peak-solar may only be available for a few hours each day, allowing for only the bulk cycle. However, without fully charging the battery in the absorb stage, both immediate capacity and lifetime cycles are reduced. This is another good reason to throw more cheaper solar panels at the problem, giving a longer charging day by increasing tail current, and maximizing battery life.
Absorbent Glass Mat (AGM)
Absorbent glass mat batteries (AGM), sometimes called absorbed glass mat, were invented to provide a more durable battery that can endure mechanical shocks and various mounting angles for motorsports use. None of these properties are relevant to an off-grid storage system, but result in these batteries costing about twice as much as the wet cell equivalents. As an example, the Duracell GC2, a 6 volt, 215 amp-hour battery, costs $84.52, while the equivalent AGM version, the GC2AGM, a 6 volt, 190 amp-hour battery, costs $179.73, a bit more than twice the price for slightly less capacity.
However, one property of an AGM battery, because of the slightly different chemistry used for durability, is that it can be recharged faster than a normal, wet cell battery. This allows bypassing of some of the worst aspects of the bulk-absorb-float problem noted above, resulting in a longer life span for the array. This increased lifetime can be twice as long as with wet cells, making up somewhat for the higher price.
Gel cells, such as the small sealed lead acid batteries used in a typical computer uninterruptible power supply, have many of the same advantages as AGM batteries. However, gel cells are so much more expensive than normal wet-cell batteries, up to five times as expensive, that they have already been lapped by more advanced battery technologies. We won’t discuss them further here.
Amp Hour Ratings
For off-grid battery arrays, this rating is probably the most important, but unfortunately, it can also be the most confusing at first. You will typically see several amp hour ratings. For off-grid use, the two most important are the 20 amp hour rate and the 5 amp hour rate. This “rate” is poorly named, but what x amp hour rate means is the number of amp hours the battery can provide, if it is uniformly discharged over x hours. Maybe it should be called amp hour capacity at the 20 hour rate, instead.
For comparing apples to apples, always look for the 20 hour rate. Some batteries are even marketed with 100 hour rates to artificially bump up their perceived energy storage capacity.
As an example, the GC2 golf cart battery is described with the following amp-hour ratings: a) 20 amp hour rate: 215, and b) 5 amp hour rate: 157. This means that if discharged over 20 hours, the battery can supply 215 amp-hours, but if discharged over 5 hours, it can only supply 157 amp-hours. Doing the math, this results in a current draw of 10.75 amps and 31.4 amps, respectively. Draw three times as much current, and the battery lasts only one fourth the time (at higher loads, the effect is progressively much worse).
Applying these numbers to a 24 volt string gives a consistent wattage of 258 watts (over 20 hours) and 754 watts (over 5 hours), respectively. You should heavily derate both of the figures as we will see in a later section.
As another example, the 29HM deep cycle marine battery is described with a 20 amp hour rate of 105. This results in a 20-hour current draw of 5.25 amps, and a consistent wattage (with a 24 volt array) of only 126 watts. Again, this should be heavily derated.
Note that in the example above, we used four GC2 6 volt batteries to make a 24 volt string, while we only use two 29HM 12 volt batteries. To be a fair comparison, we should combine two of the 29HM strings (for a total of four batteries), resulting in 210 amp hours (at the 20 hour rate), 10.5 amps at that rate, and 252 watts at that rate, very similar to the four GC2 batteries.
You will also often find the number of minutes the battery can last at a given amperage. This value is related to the above in an obscure way, but this version is at least a little easier to understand. For example, the GC2 golf cart battery is described with the following ratings: 395 minutes of operation at 25 amps, and 105 minutes at 75 amps. On the other hand, the 29HM deep cycle battery is described with 225 minutes at 23 amps, and 185 minutes at 25 amps. Note that the chosen minutes and amps are almost random. While this rating is easier to understand, it is almost useless for comparing different batteries. Stick with the 20 amp hour rating across all batteries, where possible.
This is another almost meaningless parameter when it comes to off-grid battery arrays. This value is basically how much extra juice is left if you want to completely shred the battery internals and not use them again. We will ignore this parameter.
We will discuss battery maintenance, including topping off acid, replenishing water, testing and safety in a future article. An off-grid battery array is definitely not a zero-maintenance proposition.
Capacity Versus Lifetime
You are unlikely to find useful information about how long you can expect a battery array to last in typical use, whatever that means. We plan for a battery array to last about two years, with proper maintenance, discharging no more than 50%, a discharge rate under the 20 amp hour rating, and fully float charging. We’ve exceeded these limits from time to time, such as during the Hurricane Matthew experience, but we also don’t count on the batteries 24/7, either. We also keep our batteries topped off continually rather than forgetting about them in the corner of a garage until we need them.
As with solar panels, always compose individual strings using the same type of batteries. If possible, design the entire array with a single type of battery, purchased at the same time, and with similar date codes, but the array can be composed of different types of batteries if needed, with some restrictions so long as each string is consistent internally.
The main reason each string should be internally consistent is that when charging or discharging, a mismatch would cause some batteries to be over- or under-charged. Lead acid batteries are pretty good about self-balancing as long as the batteries are similar. One weak, older battery in the string will wear all the batteries out faster. It would be better to group the weak batteries in their own string. Even better would be to dedicate those batteries to a rarely used, separate application.
Energy In Versus Energy Out
Even with careful planning, charging, floating and current management, you will still find that the batteries will only deliver about 70% of the anticipated energy. This discrepancy comes from a variety of sources, but expect this up front in your planning, and be pleasantly surprised when more is available. As a result, derate your battery capacity not only by the recommended 50% discharge, but also by this 70% fudge factor. In other words, plan to draw only 35% of the battery array’s rated capacity. For a single GC2, with a 215 amp-hour capacity (20 hour rate), expect to only draw 75 amp hours, or 1806 watt-hours for a four-battery string. Over 20 hours, this means that this battery string could supply a 90 watt load. This would be enough for a single efficient modern freezer (the remaining four hours in the day coming off the solar panels directly).
Discharge Versus Voltage
So what does 50% discharge mean, anyway? We can measure the state of charge with either a hydrometer, which tells us the density of the electrolyte and thus how much charge remains available, or with a voltmeter. The hydrometer approach is reserved for the future article on maintenance. The voltmeter approach can also produce good results more safely and easily.
For a healthy battery, we can determine the actual state of discharge by measuring the resting voltage of the battery. The battery is considered at rest one minute after a 10 to 15 amp load. The application of a recent load for a few minutes removes what is known as a surface charge, which can provide a misleadingly high estimate of the actual charge. The following table, taken from Interstate Batteries’ Marine/RV Battery Maintenance Guide, shows the battery voltage versus state of charge for various combinations of batteries:
|State of Charge
|100%||6.38 v||12.75 v||25.5 v||51.0 v|
|75%||6.23 v||12.45 v||24.9 v||49.8 v|
|50%||6.13 v||12.25 v||24.5 v||49.0 v|
|25%||6.03 v||12.05 v||24.1 v||48.2 v|
|0%||5.95 v||11.90 v||23.8 v||47.6 v|
Measuring this resting voltage requires disconnecting the load, which may be inconvenient in actual use. Because of internal series resistance, the effect of which varies with given loads and condition of the batteries, it is a good idea to apply some typical loads to your array, measure the voltage under load, remove the load, rest for one minute, and check the resting voltage. In our case, we determined that with our load applied, the batteries were at about 50% state of charge when the array under load was about 24.0 volts. After resting with the load disconnected, the batteries would show about 24.5 volts.
Now that we’ve covered the essentials of lead acid battery arrays, we can now proceed to the remaining portions of our discussion about off-grid energy. In the meantime, check out SoftBaugh’s helpful battery array calculator, which takes a lot of the guess-work out of some of the calculations. Note that the calculator shows a 50% recommended depth of discharge. For our purposes here, adjust that number to 35% when doing your planning.
A reader sends a link to an excellent battery maintenance page. The voltage values shown on that page are a little different than the values from Interstate Batteries, but this could be the difference in where the line is drawn between reserve capacity and not.