This is one in a series of articles about smart spending in radio operations and managing on a budget.
Cris Alexander, CPBE, has been director of engineering of Crawford Broadcasting Co. for 41 years. He is a Radio World contributor and past technical editor of RW Engineering Extra. He is also managing partner of Au Contraire Software Ltd., which produces AM/FM allocation and design software and database services.
Radio World: Cris, our premise is that engineers may not be able to control revenue or consumer tastes, but they can control how efficiently they operate their plants. What’s your approach?
Cris Alexander: The trick here is finding the point of diminishing returns and staying just under that point.
This exercise takes many forms and differs between studio, transmitter and other parts of the broadcast infrastructure. It is also very much dependent on market, format and situation.
In some of our markets, for example, we invest in state-of-the-art off-site auxiliary transmitter facilities because the situation is super competitive and airtime is that valuable. In other locations, we take a more measured approach and use the former main transmitter as the aux with a dedicated aux antenna on the main and only tower.
RW: How does your budget process work?
Alexander: We start with a full capital assets inventory update every August and September. We maintain this inventory in a relational database, and each item has an asset ID number with a sticker affixed to the asset identifying it as property of Crawford Broadcasting with the asset’s ID number.
Within the inventory is a table of equipment that contains replacement cost and useful life for each make/model. The information from the inventory database is, for each market, linked and pulled into an Excel spreadsheet that lists all the items in the cap rotation. Five-year columns are provided, and dollar amounts appear in the appropriate year column for each item that will come up for replacement that year — the year purchased plus the useful life.
That provides us with the subsequent year’s scheduled replacement budget plus projections for the four years that follow.
Once that is done, we send that schedule to our market engineering managers so that they know what will automatically be in the budget for the coming year.
They’re given a window of time, usually a month, to ascertain other capital needs for the year. They then submit their list of requests, which include item name, cost and a full justification.
I look at each request and may approve it, request more information or deny it based on any number of factors. I prepare a draft budget that includes scheduled items plus any approved scheduled items, and I send that draft budget to the CFO and president.
We meet, usually in November or December, and go through the budget for each market line by line. I come away with a final budget that is published to our engineering managers.
RW: How have budget pressures changed in radio in the last 10 years?
Alexander: Declining revenue in the radio industry overall has definitely had an impact. We look much more critically at budget requests than in years past. Our engineering managers understand this and avoid asking for things that are not really necessary.
In years past we might purchase a piece of test equipment or whatever that would be nice to have and make our work easier. Today, we would be more likely to borrow a piece of test equipment from another market that already has it.
Also, we’ve taken a hard look at useful life figures for equipment and extended them. We used to place a 15-year U/L on transmitters, mostly because of replacement parts availability, or lack thereof beyond that point. Now we use 20 years, and we may purchase some spares of critical items like power supplies and keep them on the shelf.
RW: What discussions do you have with your team about efficiencies in operating the plant?
Alexander: We watch and track electricity usage closely, paying particular attention to kilowatt-hours and demand.
Whenever there’s a jump from the same month in prior years, I notify the engineering manager in the market and ask him or her to account for that. Sometimes we find that a photocell on a tower is broken and the tower lights are running 24 hours a day. Other times we have found a stuck relay in an HVAC system causing the heat strips to be energized even while in the cooling mode.
There’s almost always a ready explanation for a jump in utility usage, and we seek to immediately correct the issue when we become aware of it.
Utility usage is our No. 1 expense, running from a few hundred dollars a month at low-power sites to over $10,000 at high-power sites. Running the aux into the dummy load for a period of time while the main is on the air at full power can result in an increase of $1,000 or more in the utility bill, so we’re careful not to do that.
At high-power AM sites, we employ MDCL, which saves us one or two thousand dollars per month. If we have to operate on the aux and it doesn’t have MDCL capability, we operate it at 50% power so that we don’t kick that demand meter.
RW: How do you decide if it’s best to employ a full-time engineer vs. a contractor?
Alexander: Sometimes it’s a function of who’s available. There may be no available full-time qualified engineers in an area, and the cost of moving someone from out of market is prohibitive.
Other times we may not have enough work to justify a full-time engineer. And still other times it can be determined by what resources we have in place.
In one of our West Coast markets, for example, we have a very sharp operations manager who is not an engineer but can easily take care of the AoIP studio infrastructure himself. That leaves just the transmitter site, at which he also does some site maintenance work, and we employ a contract engineer for the transmitter site only. That works out great.
In most other markets, we have an engineering manager and one or two other engineers to take care of the cluster. In those larger clusters there is no way to get by with a contract engineer.
RW: How would you go about estimating power cost savings of, say, replacing an older transmitter, or deciding between air and liquid cooling?
Alexander: That is a fairly easy equation. The straight AC power consumption of a transmitter for a given TPO is a known figure. The difference between the AC power input and RF output can be assumed to be lost as heat into the transmitter room or building, and HVAC must be designed and sized to deal with that on a BTU/hour basis.
If you can move the waste heat dissipation outside, as you can do with a liquid-cooled system, that has a big impact on HVAC size and BTU/hour figures. A liquid-cooled transmitter requires some additional infrastructure, which impacts the capital cost of the transmitter. It would be a fairly easy exercise to weigh the added capital cost against the power savings over the useful life of the transmitter and come up with an answer as to whether it would be worthwhile to go with liquid cooling.
But there are complicating factors. For example, if your auxiliary transmitter is or will be your former main air-cooled transmitter, you will have to keep sufficient cooling in place to deal with its waste heat. That results in inefficiencies that can be hard to quantify.
An oversized HVAC system will not be as efficient as a properly sized system, and it will short-cycle, which results in greater wear and tear on the compressors and fan motors as well as thermal stresses on equipment. That would, in my estimation, be a significant offsetting factor against the power savings of a liquid-cooled transmitter with outdoor heat exchangers.
RW: With many broadcasters selling towers, more stations presumably lease. How do they keep costs from getting out of control over time?
Alexander: We haven’t gone that route and continue to own most of our sites, but clearly the key is in negotiating the lease at the time of divestiture and taking the long view.
That means capping increases at a certain low value and minimizing common-area maintenance (CAM) liability. A 3% capped escalation will result in almost a 35% increase over 10 years, and that’s significant. Tower owners like to spread the love with their tenants when it comes to tower and site maintenance, and this can result in annual CAM charges that equal a month or two of rent.
All that can be dealt with when the deal is made to sell the tower. It’s better in many cases to take a fixed reduction in sale price than to have an open-ended escalation in annual rent/CAM charges. Those are the gifts that keep on giving.
RW: Do stations change signal patterns with the idea of containing power costs?
Alexander: I haven’t heard about anyone doing that, but it does make sense.
It’s not a difficult exercise to determine the point of diminishing returns with respect to population coverage. As broadcasters, we’re conditioned to produce the biggest coverage footprint that we can, but in this day and age, that doesn’t make a lot of sense in many situations.
Consider a 100 kW station serving a city. Run the population within the protected contour and see how that compares with the city population. Now reduce the power to 50 kW and run the pop again. Not much change? Drop it to 25 kW and repeat.
You can do this and find that point where the population served really starts to drop off. Your maximum efficiency in terms of power transmitted and population served is just north of that point. The reality is that anything more is mostly cows and chickens, and you’re wasting capital and operating costs to provide them with a listenable signal.
Of course there may be other factors, such as needing more power for building penetration in built-up areas, and that would need to come into the equation. But in many cases, particularly in small and medium markets, less power and coverage may translate to more bottom-line profit.
RW: What practices can you suggest for a maintenance program and management of equipment lifecycles?
Alexander: A lot of this has to come from experience and empirical data. How long does a transmitter remain reliable? How about a file server or computer? Are there other influencing factors that can drive that useful life number? Over my 48 years in this business, I’ve learned to guesstimate U/L, though sometimes I miss badly!
I mentioned that parts availability affects transmitter U/L and how you can extend that by preemptively purchasing certain critical spares, such as power supplies.
Vendor switching power supplies, popular in just about every piece of broadcast gear these days, are the No. 1 item that will fail and for which there may not be a ready replacement available.
Computers and servers, I’ve found, will start acting up after six or seven years. I’ve found that the thermal compound between the CPU and heat sink many times turns to powder in that time period, which can cause systems to blue screen and permanently damage the CPU.
And of course there seems to be a finite number of read/write cycles to any mechanical or solid-state drive. So we preemptively replace computers and servers at five years. That keeps us out of trouble.
Other equipment, such as consoles or control surfaces, can be highly variable and situation/use dependent. You watch and learn.
RW: How can you extend the life of older transmitters while still meeting compliance?
Alexander: The No. 1 issue I’ve seen with older tube-PA transmitters in auxiliary service is caused by moisture. The exhaust is almost always ducted outside, which means the stack and PA tube are exposed to outside air and humidity. The exterior of that ductwork is exposed to conditioned air.
The result is often condensation in the stack and PA area, with predictable results. Adding motorized louvers to close off the exhaust duct from the outside environment will help, as will insulating the exposed ductwork indoors. Also, running the transmitter periodically will help de-gas the tube and show up any issues.
Older transmitters in regular on-air service need to be kept clean, inside and out. Change filters regularly and make sure they’re breathing clean air. Keep an eye on stack temperatures and temperature differentials.
And keep critical spares on hand. One of the biggest challenges these days is PA tubes. Rebuilds are scarce and expensive, and new tubes are ridiculously expensive, upwards of $5,000 for a new 4CX15,000A.
We do keep corporate-level spares, mostly of power modules, but each station or site has a full set of recommended spares purchased with the transmitter. A key factor is replacing spares as they are used without fail. And we preemptively purchase vendor power supplies, not from the transmitter manufacturer, so that we’re ready when (not if!) the originals fail.
RW: What else should we know?
Alexander: We have had to change our thinking. In the past, we might well have made purchases and acquisitions on a “cool to have” basis, and we had revenues to cover that. We might have occupied “Class A” space in prestigious leaseholds.
But things have changed and continue to change. It’s a brave new world, and we have got to change our mindsets to adapt. This is true at every level, but especially at the management and corporate levels. It’s all about efficiency. We may not be able to increase revenues, but we can reduce expenses, at least to a point, to maintain a healthy bottom line. That takes effort, creativity and different thinking.
Read more on this topic in the ebook “Radio Operations on a Budget.”
