Image used under a CC license. Flicker.com/Mike Baker
Have you ever had to generate your own ac power for more than say, 24 hours? If so, then likely you know how difficult it can be. Access to utility power is an aspect of our work that most of us completely take for granted; and, for just about every studio or transmitter site we would build, there is an automatic assumption that utility power is available, and rightly so.
What happens, though, when you need to build a remote site that is so far off-the-grid that there is no practical (meaning economical) means by which you can get utility power? Does one simply write off the potential site, even though it might be fantastic in every other way? That would certainly be “letting the tail wag the dog.” Let''s look at means of generating your own power — all the time.
High in the mountains of southern California, this PV array provides much of the energy needed for KLRD''s transmitter facility.
The Economics of Generating Power
Before we get into the subject in detail, it''s clear that there is a practical limit to power generation at a remote site. Whether or not you decide to build your own “power plant” is going to depend on several factors:
1. How much power is really needed at the site?
2. How much will it cost, in both up-front and on-going expense?
3. How much will it cost to get utility power installed at the site?
Clearly the answer is going to vary with each and every site and circumstance. The first question you will need to calculate; the third is going to depend upon the particulars of the site; and we can address the second question at least in terms of the items that need to be considered as you calculate costs.
- continued on page 2
Clearing snow from the PV array in winter; just one more day in the life of a broadcast engineer.
Dual diesel generators. Rarely have I seen a mountain-top or remote transmitter site without a diesel generator as a backup power source. Even if you were to pay your local utility company upfront to run the power lines needed up the road to the transmitter site, it''s very likely that you''ll still want to put a generator in as well. At what point does it make just as much sense to install two generators, and to skip the utility power? To determine that, you must, at the very least, consider the following:
1. Actual capital cost of a second generator
2. Installation expense of a second generator
3. Cost of installing large fuel storage tank(s)
4. Expected lifetime of generator running 50 percent of the time
5. Fuel expense to run one of the two generators continuously
6. Likelihood that fuel deliveries will be difficult in bad weather
7. Generator maintenance
Dc controllers and 120Vac sine wave inverters for KLRD''s hybrid power system
I do know of one circumstance in which a dual generator system was built and put in to full-time use, because although it was anticipated that utility power would eventually be installed, it was delayed, and the station owner didn''t want to wait for it before turning up the site. John Burger was the chief engineer for KTRB during this undertaking. He described for me what happened. “In 2006 KTRB-AM was a C-O-L change from Modesto, CA to San Francisco. In SF it was designed to operate from separate day (non-D, 50kW) and night (directional, 50kW) sites. The daytime site was to be located in Sonoma County. The nighttime site is located in Sunol, California,” John said. “The Sonoma site never came to be. This consumed much time and money. Construction on the nighttime site began July 2006. The site is several miles from public roads and PG&E (the major local utility). PG&E service would have been available if access could have been secured from one of the surrounding ranches, but this was not immediately forthcoming. Therefore the site was designed and built to run on generators.
“The power plant was designed to use alternating sets of day/night generators. We had two 35kW generators for daytime (non-transmitting) operation of the site, and two 260kW generators for nighttime on-air operation. The system was to be controlled by a series of transfer switches that alternated day/night operation as well as alternate day operation. Daytime/nighttime operating hours were to be controlled by a timer. Access to the site was along a road belonging to the San Francisco Water District, which ran along one side of a reservoir that is a large part of San Francisco''s drinking water system. Due to concerns about diesel being spilled into the reservoir we were ‘encouraged'' to install LPG generators. Fuel storage is a 10,000 gallon tank. The Bay Area Air Quality Management District mandated that each large generator be equipped with a catalytic converter the size of a VW Beetle, and costing as much as a Porsche Boxster.
- continued on page 3
“KTRB''s nighttime site went live in February 2007. Because the daytime site was still being delayed by Sonoma County, the nighttime site was used 24/7. The transfer switch system was modified to just alternate the two large generators on a daily basis. The system consumed about 4,800 gallons of LPG fuel per week. It ran this way until mid-2010 when both 260kW generators reached the effective end of their lives. Under the original plan the generators should have been good for about 7.5 years before needing rebuild or replacement.
“What does this all mean in dollars? Well, the generators, transfer switches, and fuel system cost about $600,000 to buy and install. Maintenance and fueling over the 3.25 years of operation amounted to about another $1M. Several studies were done to bring PG&E service to the site. Following the most direct route, building on a pole system would cost about $1M. The longer, trenching route for buried cable, following ranch roads would cost about $1.7M.
“Running a 50kW transmitter site utilizing generators as prime power is madness, even if forced by circumstances.”
KTRB eventually abandoned the original nighttime site and now operates from a diplexed site in Hayward, CA. At the risk of over-generalizing, I''m going to say in the majority of circumstances it makes more sense to pay a utility to install power at a remote site instead of using two generators.
Hybrid systems approach. For lower-power applications (low-power transmitter, or repeater site) it can be quite practical to build your own power plant. The “hybrid” approach means that you will not rely on any single source for all of your power needs. Whereas you might base your system on solar photo-voltaic cells (PV), you would likely also have a small generator, or perhaps even a wind turbine, to back up your system. There are practical limits to systems such as this as well. Let''s examine the size of the PV array, and battery array, for a hypothetical system. (This is not meant to be a primer; consult a design engineer that specializes in hybrid systems should you decide to build one.)
Continued on page 4
Figure 1: Basic PV System
Take a look at Figure 1, which is a basic PV system. The solar array is configured to charge large batteries; dc to ac conversion (by way of an inverter) is then used to generate the actual 60 Hz/120Vac to energize the load. Figure 2 shows an auxiliary energy source applied in parallel with the PV array (thus making it a hybrid system).
So, let''s say you want to build a remote mountain-top transmitter site, which is completely off of the grid, with no practical access to utility power. The following items will be in constant use:
> New FM transmitter, with 1kW of TPO
> Main and backup STL receivers
> Main and backup audio processors
> ISM-band radio (provides duplex IP connectivity)
> One Layer-2 (Ethernet) switch
> Main and backup remote controls
> Four 26W CFL lamps
The very first part of the process is determining what the total ac power load will be. In order of usage, we have the transmitter, for which I will use an ac to RF efficiency of 60 percent, thus requiring 1667W; the four CFLs require 104W; the audio processors, each 50W; the STL receivers, each 30W; the remote controls, each 30W; the Layer-2 switch, and the ISM-band radio, each 10W. The grand total is just over 2kW (which we''ll use for our calculations).
Through every aspect of the following calculations, you will note there are always efficiency factors to consider. The first one is called “round-trip” efficiency, which is basically the amount of energy recovered divided by the amount of energy stored in a device, such as a battery. A typical round-trip efficiency is 80 percent (we will only recover 80 percent of the energy stored in our batteries in this particular case). So, before calculating the size of the PV system or batteries necessary, we take our 2kW constant load and divide by 0.8, giving us 2,500 going forward.
Rule-of-thumb is that for power levels between 1kW and 3kW, the battery systems operate at 24Vdc. (Below 1kW, 12Vdc is used; above 3kW, 48Vdc is used.) The PV array must match that, so we''ll be looking at 24V solar arrays.
Continued on page 5
Figure 2: Basic Hybrid System
Solar arrays are specified by their peak power output for “standard test conditions,” which refers to 1kW per square meter of irradiance; 25 degree C ambient temperature; and 1.5 times normal atmospheric pressure. Naturally you can assume that in practice you won''t get this much power out of the solar panel — and you''d be right. There are three factors that bring down the total power available per panel: manufacturing tolerance (subtract 5 percent); dirt, because as it collects on the array, the efficiency drops (typically 5 percent); and, since we''ll be powering a load with ac, we need to account for the efficiency of the inverter (typically 93 percent). So, for a panel “rated” at 200W of output under standard test conditions, we have: 200*(.95)*(.95)*(.93) = 168W max.
The next thing to consider is just how much sunlight you can gather each day. That of course is going to vary over the course of the year, and it''s going to be higher in sunny climes like the U.S. southwest. You''ll need to determine what your peak sun hours (PSH) are for the location you intend to use. (One example: bigfrogmountain.com/SunHoursPerDay.html) I''m going to use southern California in this example, with its six PSH (averaged over the year) per day.
Let''s go back to our load calculation now: we needed 2.5kW for each of the 24 hours per day, which equals 60kW-hours per day. Each panel will provide (168W)*(6 PSH) = 1,008W-hours per day. Dividing the needed power by the amount that can be collected per panel gives you 59.5 panels; obviously we''d round up to 60.
Now we can begin to see how practical our theoretical hybrid system is. If you consider a panel such as a Kyocera KD205, you''ll note that the dimensions are basically 60” by 39” by 1.4”. It weighs 40 pounds; 60 such panels will weigh 2,400lbs. (plus the weight of the framework). There are any number of ways to mount the panels, but say you were to put up two rows, each 30 panels wide; you would then have a rectangle 97.5'' wide, and 10'' high. The framework will be designed in such a way as to optimize the angle of the panels with respect to the sun.
Battery selection is the next step, but before we do that, we need to know a few things. For starters, each battery will have a depth of discharge specification that we''ll need to know. (Typically this is 80 percent.) Secondly, we''ll need to decide how many days of autonomy we want — basically the number of days the system will operate without any input from the PV array. For our purposes, let''s use three. So, to figure out our amp-hour rating, we divide our watt-hour need by the battery system voltage; we divide that result by our D-o-D figure; and finally we multiply that result by the number of autonomous days: 60,000/24 = 2500; 2500(3)/.8 = 9375 AH total.
Battery selection is interesting because it would seem to make sense to put two 12V batteries in series, and then add parallel branches to build the AH rating. (Batteries added in series keep the same AH rating; but when adding parallel branches, the AH ratings add.)
Continued on page 6
Figure 3: Basic Wind Generator System
Let''s take a specific example that''s easy to find: Surrette batteries. (rollsbattery.com). In order to get the AH rating we need, you could use an array of 54 12V 357 AH batteries (27 parallel branches each made up of two batteries in series); or, by using 4V 1350 AH batteries, you can achieve the necessary AH rating by building an array of 42 batteries (seven parallel branches, each made up of six batteries in series). It turns out that the cost of the 4V array is somewhat less (on the order of 5 percent) but there are some other important factors: the 4V battery array (when flooded) weighs 13,230lbs.; the 12V array weighs 14,688lbs. The 4V batteries need 48 square feet of floor space; the 12V batteries need 92 square feet. (I''m neglecting the height differences because no matter what, the room will need to be tall enough for you — and you''re much taller than either type of battery.)
So now we know how large the PV array is, and how large the battery array is. What I want to point out is that in calculating the overall cost of a hybrid system such as this, you''ll add up the hard costs of the solar cells, and the batteries, but you also must include room indoors for the batteries, as well as the cost of whatever structure you use to hold up the solar cells.
Since we''re describing a hybrid system, we must consider the alternate power source as well. Consider a small generator first, the purpose of which is really two-fold: It can operate the entire ac load should the need arise, although normally it is used to keep the battery array from discharging too far in the event that dc input from the PV array is lacking (take a look at Figure 2 again.) Obviously, in this configuration the generator must be large enough to handle the complete ac load.
Another source of power for charging the batteries is a wind-generator. (See Figure 3 at RadioMagOnline.com) I''ve written before about using wind for power: radiomagonline.com/infrastructure/power/radio_sun_wind_fuel//index2.html. Since the efficient use of wind-generators requires the installation of another tower, and depends upon clearance around the wind-generator itself, I doubt that there will be many instances favorable to this type of system.
Another important item in the system is the controller, which is often a combination unit that not only serves as the inverter, but as the battery charger as well. In some systems, the charger/voltage regulator that feeds dc to the batteries, either from the PV panels, or a wind turbine, is added into the system.
Let''s look at a real-world example of an off-grid, hybrid energy system. KLRD is one of Educational Media Foundation''s stations in Southern California (licensed to Yucaipa, which is about 70 miles east of Los Angeles). It has an ERP of 590W, but with a HAAT of about 3,400'', and height above sea level of over 9,000'', it can be heard over a very wide area. The Mt. San Gorgonio site is very isolated, and in fact, the site''s Engineer, Jeremy Preece, tells me that access can only be gained during winter months by flying into the site. I asked Jeremy to describe the hybrid power system that KLRD has been using successfully since 2001. “The (KLRD) system was designed to comfortably support a constant load of about 500W, or 12kW-hours per day. In reality we use closer to 15kW-hours per day since we have added more ancillary equipment and increased TPO for an upgrade a couple years ago.
Continued on page 7
“The system uses (24) 4V 1,550AH flooded-cell batteries, wired in series-parallel to make a 48V system. These batteries provide a total of 2,500AH storage at 48V. The solar array uses (68) 75W panels to give us a peak input of 5kW, although this is only at prime conditions, as solar panels lose so much efficiency when not precisely aimed at the sun. The system on average is able to produce between 2kW and 3kW, most of which is pushed to the batteries. A sophisticated digital charge controller manages the flow of energy from the panels to the batteries. Due the potential weather at the site, the panels are stationary and tilted to favor winter sunlight. While possible, we do not change the tilt during summer months as it is simply too much work. 120Vac is created by a Trace 4kW pure-sine inverter and it has been going strong for 13 years now. We have a backup on-site, too. We also have a LPG-fired 8kW generator (which at 9,200'' MSL is de-rated to a mere 3kW), which is started automatically when the system voltages drop below a specific threshold.”
I asked Jeremy about the design of the battery bank — specifically why lower-voltage batteries were chosen for their system. “My understanding on why we use low-voltage batteries is mostly storage space and simplicity, but there are other considerations. Each 4V battery is about 1,500AH (6kW-hrs) so you get a fairly large bank with few batteries. By comparison, a 12V deep cycle marine battery can store about 200AH, or 2.4kw-hrs of energy. When you calculate the load banks with Ohm''s Law, you see that in order to obtain 48V at 2,500AH, you''d need about 100 12V batteries to do what you can with 24 4V batteries. That is an enormous number of inter-connections that must be cleaned and maintained, plus the regular ‘watering'' of each cell. In a 12V battery, that''s six cells per battery, so you have a lot of work to keep them all topped off. The plates inside each low-voltage flooded battery are designed for more stressful cyclical use. They are less prone to pitting from charge/discharge cycles and have more surface area to resist sulfate build-up, which is why they tend to last so much longer than small, higher voltage batteries — the warranty on the 4V batteries is 15 years, if they are properly maintained. By comparison, a good marine battery has a warranty of 2 or 3 years in deep-cycle service. So, you get a lot more bang for your buck with the big batteries.”
“Green” power is all the rage right now, and many vendors are available to sell you a system. As we''ve seen, it''s completely practical, as long as your power requirements aren''t too high. Just remember all of the design factors involved when you are determining the overall capital and operating costs of the system.
Irwin is RF engineer/project manager for Clear Channel Los Angeles. Contact him at firstname.lastname@example.org.