Fullscreen HTML5Med Res (??MB)HTML4

To make solar power workable and cost-effective, there are a few rules to be followed. Much of it is commonsense but some aspects are not obvious, like making sure that battery storage is matched to solar panel capacity. Here, we look at the best approach. Solar Power for Caravans & Motor-homes: Dispelling the Myths By COLLYN RIVERS* T falling on the more habitable parts of Australia averages 1000 watts per square metre. Only 10% of that can presently be turned into electricity but this is still enough to be useful. My off-road OKA motor-home runs a 70-litre fridge, multiple halogen lights and an Iridium satphone, all from two 80-watt modules. It has not he energy of sunlight 12  Silicon Chip run out of power in the past seven years. My all-solar-electric home north of Broome runs from an 1800W solar array and has enough energy left over each day to irrigate 150 trees. Solar energy really can be made to work but there are a few traps that can result in less energy being captured than expected, and even less ability to store and retrieve it. The most common result is that your storage batteries will run down much sooner than expected. Worse still, because they are not being fully charged, many expensive storage batteries will expire within a year. The biggest trap relates to solar module output – the industry uses the term ‘panels’ for assemblies of modules. www.siliconchip.com.au Solar modules are curious devices that only produce their claimed output in quite specific applications and ‘Standard Operating Conditions’ that bear little or no relationship to reality. Watts ain’t necessarily Watts A watt is defined as one amp multiplied by one volt. To produce 80 watts, a module feeding a system operating at say, 12.8V MUST therefore produce 6.25A. But Table 1 (which is from the back of a real-life 80-watt module), shows it only puts out 4.6A. The solar module industry is not known for understatement so you can bet that the output is not a tad more. Here’s how the arithmetic is worked out: Solar modules produce much the same current across a wide range of load voltage. To establish maximum output, the solar industry plots load voltage against current and picks whatever combination gives the highest number. Physics being as it is, for the module (Table 1) to develop 80 watts at 4.6A, that 4.6A has to be developed with 17.3V across the load. This is fine if your system runs at 17.3V. Such systems being as rare as sardines that ride unicycles, the only way you can fully utilise an output at 17.3V is via a DC-DC converter that gives more amps at less volts (these are sometimes used in sophisticated large-scale systems), or by driving a load (such as some water pumps) whose output is proportionate to input voltage. If the load is a 12V charger, the most energy transferable (for the module in Table 1) is 4.6A times (say) 14.5V, ie, about 67 watts. If the load is 12.6V, the most that can be transferred is 58 watts. Temperature losses Mono and polycrystalline modules lose about 4-5% of their output for every 10°C increase in temperature. *About the author . . . Collyn Rivers, shown here working in his all-solar home north of Broome, WA, is well known as the Founding Editor of Electronics Today International which, in 1976, was proclaimed the ‘Best Electronics Magazine in the World’ by the Union Internationale de la Presse Radiotechnique et Electronique, and was produced as separate editions in Australia, UK, Canada, France, Holland, Germany, India, and Indonesia. The rated output is measured at 25°C but this does not refer to the ambient temperature; it refers to the operating temperature of the cells. Typically, at 25°C ambient, those cells will be around 55°C (under a hot sun) so there goes 12-15% of the output. At 35°C the loss is 16-20%. In contrast, amorphous technology (Uni-Solar, Solarex Millennium) modules increase their output slightly as temperature increases. In practice, a 64-watt amorphous If you’re planning to get off the beaten track but still want a few creature comforts (like lighting, TV, computers, etc) solar power is the way to go. It’s not difficult to install and set up but there are a few pitfalls for the unwary . . . www.siliconchip.com.au July 2003  13 Table 1: an “80W” solar panel’s ratings reveal that the eighty watts is mainly a figment of the manufacturer’s imagination (or at least their marketing department’s . . .) module produces the same as an 80-watt module of any other type, once above 36-38°C. But they are about 30% larger. For any practical purpose (which does not include a 17.3V caravan system at the top of Mt Kosciusko), an 80watt solar module produces about 58 watts or a bit less, in very hot places. Most modules reveal this but only in the fine technical print. Many systems fail to deliver because someone (not unreasonably) assumed a module’s amperage is the rated output in watts, divided by about 12.0 (volts). Solar regulators Interfaced between solar modules and the load, solar regulators ensure that batteries charge as rapidly and efficiently as possible. They also maintain the system at approximately 13.6V, once the batteries are close to fully charged. The most basic are voltage-sensitive on/off switches. The more complex use pulse-width modulation and incorporate all-but-essential system and battery monitoring (see below). A solar regulator should be used in every system, except where solar output is less than 0.5% of battery capacity. Peak Sun Hour contours for July (above) and January (below). Multiplying true module output by the relevant number of peak sun hours gives the wattage output for one day. There is no need to correct for changes as the sun moves across the sky. These (redrawn) maps are based on Australian Bureau of Meteorology data. (Taken from “Solar That Really Works!” by the author.) Beware of ‘self-regulating’ modules. These have insufficient voltage to overcharge a battery and in hot places their temperature loss may be such that they will not charge a battery at all. Battery traps Ironically, some of the worst people to ask about batteries are those who work in general electronic disciplines! The (US) Ample Power company states that, [to understand batteries] “general electronic knowledge isn’t enough... even those working in battery distribution channels can’t be relied upon to dispense correct and meaningful information”. Deep-cycle batteries in particular are complex mechanisms. A short article like this cannot make you an expert but hopefully it covers the essentials – and may show how some of you are killing batteries right now. All lead-acid batteries have internal resistance. That internal resistance is described in ‘Peukert’s Law’ (for14  Silicon Chip www.siliconchip.com.au This limited charging of car batteries is not a problem for starting. The starter motor is designed to work at the corresponding voltage. Limiting charging to 14.2-14.4V also safeguards electrical components. The car battery’s only major role (apart a voltage reference) is to start the car. If you want to win bets, ask your friends how much energy this needs. The answer usually surprises most people – it’s negligible. The starter motor gobbles 300-400A but typically for less than five seconds. This is about 0.5Ah or what a tail-light draws in about 15 minutes. The alternator replaces this in a minute or two, by which time the battery is back up to about 65% charge. But from there on the charge rate tapers rapidly. By 70%, charging has dropped to an amp or two and is falling fast. The battery still continues charging but very slowly. Given long enough it will eventually over-charge but that takes hundreds of hours. For most vehicles, battery charging effectively stops at 70%. Disaster for house batteries mulated in 1897) which states that the greater the rate of discharge, the greater the internal loss, hence the lower the percentage of charged capacity that can be used. It’s like the inverse of pouring beer quickly into a cold glass – the quicker you pour, the greater the foam and the less the glass is filled. You may want to repeat this experiment a few times (hic). A battery is charged by applying a voltage across it greater than it already has. The charging rate is more or less proportional to that voltage difference, so it tapers off as the battery gains charge. Constant voltage charging If the charging voltage is fixed, then as the battery voltage rises, the charge rate automatically falls. This is how a car alternator/regulator works. It’s called ‘constant voltage charging’. When used in a car system, it does not and cannot fully charge the battery. It’s deliberately designed not to. Some vehicles are driven for many hours a day (like taxis on shift work) so it’s necessary to prevent overcharging. This is achieved by limiting charge voltage to 14.2-14.4V. This corresponds to about 70% of nominal battery capacity, after which the charge rate rapidly tapers off. The battery continues to charge but so slowly that it takes 100 hours or so of non-stop driving to even approach full charge. If charged at that voltage continuously however, the battery will eventually be over-charged. The charge voltage is therefore very much a compromise. Battery makers say that, with caravans and motor-homes, 65% of full charge is typical and 70% is rare. www.siliconchip.com.au This charging regime is OK for the starter battery but far from satisfactory if used to parallel-charge a ‘house’ battery in a caravan or motor-home, not just because of the 70% or so limitation but also because the extra alternator capacity needed to achieve that in reasonable time is unlikely to exist. This can be a problem as it will also affect the starter battery in the same way. Even the best batteries are progressively damaged if they are frequently discharged below 50% capacity. This then leaves a mere 20% of battery capacity available, if one follows their makers’ advice. In practice, most people discharge their batteries until the fridge stops working, which corresponds to about 80% discharge. Even discharged this deeply, only 45-50Ah can be pulled out of a 300Ah battery charged to 65%-70%. And each time you do it, 0.5% of the remaining battery capacity goes to sulphate heaven. There are various ways around this. One is to use a ‘smart regulator’. Alternator willing, these initially charge at a constant current of up to 25% of battery Ah capacity. Once past 14.4V or so, charging is cut back to about 10% of Ah capacity to allow the charge to be absorbed. This is usually followed by a ‘float’ level of about 13.6V. There are several really good smart regulators now available in Australia. Another solution is to accept the limitations of the charging system and switch to gel cell or AGM batteries. Table 2: typical daily power requirement for a medium-sized caravan. Of course, individuals may vary significantly from these figures but they give you an idea of where to start with your own power requirements. Add a microwave oven and you’ll blow these figures right out of the water! July 2003  15 Both charge close to 100% from only 13.8-14.1V and can be discharged more deeply than conventional batteries with less internal harm. Yet another way, adopted by many caravanners and a few motor-home owners, is not to rely on vehicle charging at all. Their house battery charges from solar alone. If you drive more than a couple of hours most days, it pays to use vehicle charging, especially if you add a smart regulator. If you don’t, it doesn’t. (Note: smart regulators cannot be used with today’s electronic engine management systems.) Battery monitoring Lead acid batteries store energy in the form of chemical reactions between lead plates and a water/acid electrolyte. These reactions are extremely slow so little is gleaned from instantaneous voltage measurements except that the meter is working. A close to ‘flat’ battery will present as close to fully charged after a few minutes on high charge – an otherwise well-charged battery will present as ‘flat’ for some time after running a microwave oven. Hydrometer readings are a little better but not much. The only meaningful indication is the voltage after the battery has rested literally for three days (and even then the error may be 15%). A very much better way is by measuring what goes in and what comes out and deducting a bit for system losses (but even this is inaccurate unless corrected for Peukert’s Law). This function, plus many others, is now built-in to most up-market solar regulators. These cost around $300 upwards. Supplementation or self-sufficiency? There are two main approaches to using solar power. They may not seem that different but the technical implications are profound, as is the effect on battery longevity. The first approach is to use solar to supplement the energy already in the battery from vehicle charging. This lets you stay longer on-site but sooner or later (and usually sooner, because you probably started at 65-70% charge), you can no longer keep the tinnies cold. All told, it is better to have sufficient solar input to be self-sufficient. This needs surprisingly little more capacity if you are setting up to stay at least 5-7 days on site. The big difference is that the first way has batteries being continually and deeply discharged – and commonly flattened. The self-sufficient way has batteries remaining close to fully charged. They typically rise beyond 95% during the day, dropping to 80% overnight. Batteries just love this, and return the compliment by living forever. And there’s no ongoing concern about the battery running down. Available energy This one’s easy. The solar industry quantify sunlight in units called ‘Peak Sun Hours – commonly abbreviated to PSH, or just ‘sun-hours’. A sun-hour is like a 50-litre drum of sunlight of uniform density: no matter where or when it is gathered, the drum contains the same amount of energy. The same people produce sun-hour maps that use contours to show the average number of sun-hours at different times of the year. Most sun-hour maps show irradiation in units that need juggling to be meaningful. The sun-hour map in this feature needs only the relevant sun-hour number to be multiplied by the (true) module output. For example, an ‘80-watt’ module (realistically 58 watts) produces 175-350Wh a day in most places one visits from choice. Cloud cover and smoke Sun hour maps allow for average cloud cover but there are likely to be exceptional days. It is extremely rare to experience zero solar input. Heavy cloud typically cuts input by 50%. The greatest loss is heavy cloud and rain and also even light smoke from bush fires. Irradiation is commonly diffuse, so light haze may actually increase it, particularly near water or light coloured sand that reflects back to the haze layer. Module orientation Over time, optimum input is obtained with the module/s facing into the sun but having the modules flat on a vehicle roof is an acceptable compromise. Except for way down south, there will typically be 15-20% loss and this is readily and cheaply compensated for by adding the equivalent solar module capacity. What can be powered Two items are typically responsible for 70% of daily electrical consumption and system cost. These are refrigerators and microwave ovens. A really efficient 40-70 litre chest-type compressor-driven 12/24V electric fridge uses 250-350Wh/day. A larger (say 110-litre) front-opening fridge of the same type uses 500-600Wh/day (Wh is watt-hours). These are realistically the largest electric fridges that are practicable Batteries for Solar Power Systems Pictured at right is the "Sungel" battery, an Australian designed and manufactured battery specifically intended for remote area power systems, including solar systems. Developed in conjunction with the CSIRO, the battery is claimed to have a 12+ year design life (double the life of other gell cells) and is available in a range of sizes and capacities. Where most lead-acid cells cash in their chips with deep discharge cycles, the Sungel is claimed to suffer no ill-effects with continual 25% discharging (5000+ cycles) and will still give 2500+ cycles at 50% discharging. Even an 80% discharge regime will still yield 1500+ cycles. The manufacturers, batteryenergy, also have an even higher-rated VRLA gell cell, the energel, with a 20+ year design life. For more information, visit www.batteryenergy.com.au or call batteryenergy on (02) 9681 3633. 16  Silicon Chip www.siliconchip.com.au to run from solar power (unless you run a solar module franchise on the side). Better by far are the three-way gas/12V/240VAC units. These run on 12V while driving (when they pull up to 15A). They can be run on 240VAC mains power if and when available, and gas at all other times – NEVER while driving. Microwave ovens are energy gobblers. Most people assume that because they may say 600-800 watts on their fronts – that’s what they draw. That rating is the heat equivalent of the energy they produce, NOT the electrical energy consumed in doing so. The latter is typically 60% more. Another 15% is lost in the big inverter needed to drive it (big sine-wave inverters drop off in efficiency at close to full load) . Driven via an inverter, these ovens typically draw 150 plus amps (at 12V). Ten minutes running a microwave oven equates to the better part of a day’s output from a 64-watt module. Apart from the above, you can run most appliances except those whose primary function is to produce or shift heat. The most efficient lighting is the still developing LED technology, followed by fluorescent (compact globes or tubes) and halogen respectively. Incandescents draw too much to consider (four times that of fluorescent lights). Sizing the system When assessing probable daily consumption, add 10% to most things driven via an inverter (15% for microwave ovens) and another 10% to everything to allow for charging/discharging losses. The total result for all your proposed appliances is typical daily usage. If it varies much from Table 2 go over it again or your system will be bigger and cost more than most. If you intend only to supplement the battery energy, calculate your proposed battery availability (from probably initial 70% charge to your decision on discharge level). The amount available is typically 30-35% of nominal Amp-hour capacity, ie, 30 amp-hours from a 100Ah battery. Divide the above by the number of days you want to stay on-site. This gives you the amount available per day. If you stay three days, you have 10Ah available. From your probable daily usage, corrected for losses, subtract the daily battery energy available. The difference is the amount of you need to produce each day. From actual module output, calculate the number of modules you need. Calculating self-sufficiency: Calculate probable daily energy (corrected for losses). Much of the information for this article comes from Collyn Rivers’ recent book, “Solar That Really Works – Caravan Edition”. It goes into the subject in significantly more detail. The book is available at $37 including postage and packing, direct from the publisher, Caravan & Motorhome Books, PO Box 3634, Broome, WA 6725. Phone 08 9192 5961 Website: www.caravanandmotorhomebooks.com www.siliconchip.com.au There isn’t much of Australia which Collyn Rivers and his wife Maarit haven’t crossed. Their WA-made OKA fully solar-equipped 4WD off-roader is seen here crossing a sand dune in the Simpson Desert. Calculate module capacity needed to provide the above plus 15%-30% (to enable rapid battery recovery following exceptional loads and cloud cover. Suitable battery capacity should not exceed five times total daily solar input, eg. two 80-watt modules typically operating with five sun-hr/day are likely to produce 116 x 5 = 580Wh/day (or a bit under 50Ah). The optimum battery capacity is therefore 250Ah but since this may weigh 100kg or more, lack of weight-carrying capacity may limit it to less. A deep-cycle battery used in a properly designed self-sufficient application can be assume to be 90% charged most of the time and the occasional deep-discharge (eg, to 20% remaining capacity) is acceptable. About 70% of nominal capacity is thus available for use and a 350Ah battery bank will be fine. Sun-hour assumptions Plotted sun-hour data is surprisingly accurate but knowing this is of no help unless you know where you are likely to be, and when. As a general guide, solar self-sufficiency is practicable from 2 sun-hours/day if you use a gas/electric fridge; from 3 sun-hours/day if you have an efficient 40-70 litre chest type fridge, and 4 sun-hours/day for a door-opening electric fridge (but this will still need a lot of modules). If you go for an electric-only fridge, it’s advisable to have back-up generator – preferably a DC unit producing up to 15V for quick battery charging. Most combination 240V AC/12V DC generators cannot produce anything like 15V and therefore will never fully charge a battery. If you design the system assuming four or more sunhours/day, it’s advisable to allow for adding further solar capacity in the future, ie, by installing adequate cable and solar regulator capacity. The Golden Rule Never have more battery capacity than you can speedily re-charge. If you need to economise, cut back on battery storage not solar modules. If you cannot generate it, SC you cannot store it anyway. July 2003  17