November 2005

The Hidden Costs of Automation
Automation in a plant can make production more efficient and profitable — but only if you know how to avert the hidden costs.

by Gary Wineland

In many cases, this has proven to be true. Personally, I believe that plant automation is great, with a caveat — the installation that goes along with it and the corresponding costs.

I’m sure that many aggregates managers can identify with me — they know exactly what I’m talking about. If we’re not careful, it can sometimes cost more to install a device than to purchase it. This defeats the idea of using automation to simplify, streamline, and make production more efficient.

Avoiding extra costs

When reviewing the big picture and cost justifications, it should be a high priority to avoid the hidden costs associated with sensor price and installation.

During the past 20 years, automation has changed very little. The most notable change is the price of the Programmable Logic Controllers, or PLCs, and computers. In the early 1980s, we purchased one of our first PLC — a Westinghouse Numa Logic.

That PLC had 24k of memory and was limited in control registers as well as the number of timers, counters, and internal relays. In today’s world, it would cost a few hundred dollars — at most. But back then, the Numa Logic was more than $10,000, and the cathode ray tube used to program it was $7,000. We still have that old CRT TV-type screen.

In addition to the equipment costs, we needed to build a panel for manual override. Touch screens were available, but the cost was astronomical.

Well, here we are in 2005, and it’s hard to believe that we can actually spend less to automate our plants now than we did 20 years ago. But this is true in regards to the PLC and computer devices that control the plants. Further, the sensors that send information to the computer also have remained low in price. This is very important when it comes to justifying automation. After all, those of us involved in this field know that we justify our employment existence by keeping our budgets low and manpower at a minimum.

On the negative side, labor costs have significantly increased. Therefore, it is important to find ways to install and maintain sensors in a manner that minimizes labor. So, how can we keep the costs associated with automation to a minimum? A few examples follow.

Complex and cumbersome

Here we have an application where you need to know if material is flowing out of the gate (refer to Photo 1). Let’s assume that we do not need to know how much but rather a flow/no-flow condition. Let’s use an old, but widely used method that consists of a limit switch with a paddle attached to it and list what we would like this sensor to do. First, we need the signal from the limit switch to end up in the control room PLC or other indicator. Also, it would be convenient to have a pilot light mounted close to the limit switch so we could adjust the paddle and see the indication at the conveyor rather than in the control room. This light is also helpful for diagnostic purposes.

Next, a plate for the limit switch to be mounted on and a paddle to mount to the limit switch should be fabricated. Then the entire assembly must be welded or bolted to the bin.

When this has been completed, wire the limit to your existing enclosure. You must then adjust the limit so that it operates with the correct amount of material depth. If you want to know if the unit is working properly, you have to install an indicating light on the enclosure. This light is going to need a neutral, as well as the 110 hot, to operate. Here’s what is needed to install this flow switch: A limit switch, fabricators for plate and paddle, and a welder and welding truck at the conveyor when the fabrication is complete. An enclosure is not needed, except if a pilot light indication is needed at the location. Remember, a neutral wire — the “white” wire that is needed for 110-volt AC operation (all 110-volt AC needs this neutral to operate) — is necessary if a pilot light is needed.

As this example indicates, a lot of pre-planning is required to achieve our goal. Nothing from this application is an “out-of-the-box” installation. From experience, I know that the more planning it takes, the more money it costs, and the more things can go wrong. Keeping this in mind, let’s see what our alternatives are.

Streamlined and simple

Here is the same conveyor and same application (refer to Photo 2). This is a diffuse photoelectric sensor. It is a two-wire, 110-volt A/C adjustable sensing distance, and it has a built-in light. It is shown mounted with a simple magnetic holder and bracket. Simply attach the photo sensor to the magnetic mount and connect to power. When completed, you simply put some material directly under the photo eye and adjust the screw until the light comes on. A photoelectric sensor and mounting hardware are needed, but this eliminates the need for a fabricator.

Not much is needed in the way of planning. It’s a simple “out-of-the-box” installation, but the end result is the same. The phrase, “Thinking outside the box,” takes on a new meaning when you purchase a product that can be installed in this manner.

Plan ahead

Picture this scenario: It’s Christmas morning, and you watch as your child or grandchild excitedly opens a present. You are excited because you realize that once the child opens that special present, he are she will be thrilled. The child has been talking about the particular toy for months.

The moment arrives. You watch as he or she franticly rips the paper from the box, opens the present, hands it to you and says, “Can you make this work?” The child stands in front of you with big eyes and a wide grin. You look at the box and see the three dreaded words — Batteries not included.

How does this relate to automation? The comparison is when you purchase a particular sensor that needs low voltage to operate, and you don’t realize it until the day you arranged the installation. Usually this low voltage is DC, but in some cases it may be a low AC voltage, such as 24 volts. I won’t dwell on this, but you need to remember that when deciding on a sensor, always ask your supplier about the necessary voltage.

Research before you purchase

Let’s compare two proximity sensors. They both are similar in price, but one uses 20 to 220-volt AC or DC for power and the other uses only a 24-volt DC. A good DC power supply costs about $100. Again, out-of-the-box is not necessarily the case. Further, you now need to run an isolated conduit or shielded wire from the DC power supply to the sensor.

Here is an ultrasonic sensor (refer to Photo 3). It seems simple in design. You hand it to Bill, your electrician, with the instructions and send him on his way to the jobsite that is 75 miles down the road. You’re having a pretty good day until the phone rings. It’s Bill. He tells you he has it mounted, wired, and is ready to calibrate the sensor, but he doesn’t have the plug-in calibrator. A feeling of emptiness comes over you as you ask Bill, “What plug-in calibrator?”

He explains that the manual shows that you need a calibrator that plugs into the unit. You call your supplier and ask about the calibrator. He tells you he forgot to ask whether you needed it so you order it, and call Bill and tell him to come back. Two days later, you get the calibrator and send Bill back down 75 miles to complete the job. Bill climbs up the 30-foot ladder to the top of the bin. He removes the cover so he can plug in the calibrator, and just then, a yellow jacket starts buzzing around him. He swats at the nasty bug and accidentally knocks the calibrator into the bin. He watches as the calibrator slowly disappears out the gate and into someone’s 8-yard load of concrete.

The moral of this story is that when you buy equipment, you should try to find out everything about it before the purchase. Further, keep it simple. Our industry is not loaded with rocket scientists that travel around in white vans with oscilloscopes, density meters, and every other testing piece of equipment known to man. Try to choose products that you can diagnose with a simple digital voltmeter and on-board calibration systems.

Avoid mismatched signals

This picture shows what happens when poor planning results in having to add interface modules because of mismatched signals (refer to Photo 4). Let’s assume that you have a PLC- or PC-based system with multiple 4- to 20-MA input cards. And let’s keep it that way. If you decide on a sensor for in the plant, make sure that it is compatible with the PLC or computer interface. If you make the mistake of purchasing a sensor that has only a 0- to 10-volt DC output, you either have to buy a new card for the host or an interface module that will convert 0- to 10-volt DC to a 4- to 20-MA signal. Most of these modules cost around $300, so the cost of the sensor has been greatly increased.

During the past few years, most sensors have been manufactured to provide either 4 to 20 MA- or DC-voltage output and are usually field selectable. In cases where two or more add-on modules are needed to correct the problem, the cost of the entire job could triple.

Keep it simple

Now that you have finally convinced your manager that variable-frequency drive, or VFD, is the correct approach in your plan to vary the speed of the feeder conveyor, the question is what brand to buy. It’s similar to buying a dishwasher or clothes dryer. Think about it. You’re at an appliance store looking at two dryers. The one has a timer and a start button. The other has every imaginable function, including a built-in cell phone that will call the dealer when a problem exists. Well, maybe it’s not quite that sophisticated, but it still contains many features that are not really necessary to simply dry your clothes. Being a high-tech kind of guy and someone who used to think that the more features something had, the better it was, I was brought back to reality on a shopping trip. My wife and I were shopping for a dryer. All the extra features looked interesting to me. But my wife is the one who does the laundry, so I turned the decision over to her. Without even hesitating, she clearly informed me that she wanted the bare bones model. Curiously, I asked, “Why?” She informed me that she didn’t want to be confused when she was drying clothes. She wanted to be able to simply put the clothes in the dryer and to turn the timer knob. Further, she explained that she didn’t need some onboard computer to make the decision. After all, she’s been doing this for 40 years, and she knows what works.

The same is often true in automation of an aggregates plant. When looking to add automation, be sure to thoroughly examine the application. Keep in mind that once you install the drive, you will also have to service it at a later date. Nothing, and I mean absolutely nothing is more frustrating to the maintenance personnel when a situation, such as the following, occurs.

Let’s set the stage. It’s Tuesday morning. Joe, the electrician, arrives at work a little tired. He watched the entire Monday night football game, complete with two overtimes. He receives a call from George, the plant operator, who says that the VFD on crusher No. 2 isn’t working. Joe also learns that the plant manager is on location so he climbs in his van and heads out, wondering what the problem is. He enters the switch room to find the plant manager, George, and the rest of the guys staring at the blinking display on the drive. Joe makes his way through the crowd to the drive, feeling a rush of anxiety similar to being lost in the mountains or woods during a hunting trip. The entire group is standing around him. He opens the drive cabinet and reaches for the 2-inch thick operator’s manual. He starts to sweat profusely as he wonders what to do next. He looks at the blinking display, which is flashing, Error code 1123. Trying to look like he knows what he is doing, he desperately searches for a cure to Error code 1123. The manual might as well have been written in a foreign language. There is no time to think, and everyone is asking how long it will take to fix. Frustration — and possibly panic — sets in.

If any of you have ever experienced this, you know where I am going. Keep it simple! Don’t buy something you don’t need. Further, if your maintenance personnel have the choice of studying a manual that is either as short as a comic book or as long as War and Peace, I think you know what their choice would be. Most of the VFDs used on your plants only need a few options to complete the job. The most important of these options are a manual override button and a knob to adjust the speed. Others might include process followers that will accept an analog signal from the computer, adjustable torque settings, and a few others. The vast majority of the time, these options will be sufficient for the application. But don’t let this example take anything away from the more exotic VFDs on the market. In some cases, the advantages of these drives that will do just but anything but make your breakfast will solve some of the most involved problems associated with automation. But keep in mind: the more complex the drive, the more knowledge and skill needed to maintain the system.

Time equals money

Everyone has multiple conveyors in their plants. The PLC or computer-based control system needs to know if that conveyor is actually running. There are numerous methods by which the sensor can be mounted onto the conveyor. This picture (refer to Photo 5) shows a tab welded to the conveyor shaft. A proximity switch is then mounted close to the tab so it can sense the tab as the shaft is rotated. As you can well imagine, this is not the most efficient method of installation, not to speak of disassembly if the bearing needs to be replaced. In this example, there are several costs involved — sensor cost, fabrication of guard and plate, and welder costs to spot-weld the tab to shaft.

Here is another application (refer to Photo 6). As you can see, this unit has a threaded shaft. To mount this to a conveyor, you have to drill and tap the conveyor shaft. Let’s go step by step into the installation.

It’s Friday. You’ve received the sensor and holder, and it’s scheduled to be mounted the next day. You’ve tried to remember everything and everybody necessary for this installation. But looking at a job that is 20 feet in the air, you realize you needed a man lift so you hurry back to the office to call for one. The supplier says it will cost extra to get one on a Saturday. But knowing that you need it, you give the okay — estimating that it will cost more than $200.

Saturday comes, and you  send George up in the man lift to drill the shaft. He realizes he needs a 20-foot extension cord for the drill. You go back to the office, and two hours later George shows up and informs you that he broke off the drill bit in the shaft. You tell him to drill the other side and send George back to jobsite. In the meantime, the man lift runs out of fuel.

A couple of hours later, George comes back again. He tells you that he has finished drilling the shaft. You say, “Great,” and hand him the $55 tapping fee for the hole he just drilled as well as two cans of spray lube.

You see George coming back to the office again a few minutes later. He now informs you that the hole in the shaft was tapped. “Time to call the electricians,” you say to yourself. However, it’s lunchtime when you call, and they say that they will be over later. Much later, the electricians finally arrive and look at the job. The splice box for the sensor connection was on the side where George broke the tap off. The electricians inform you that they needed to get some liquidtite cable and run it under the conveyor to get to the other side. After all of this, the job has finally been completed by very late afternoon on Saturday.

But now let’s summarize the approximate cost of this job: Cost of unit, $245; labor to drill and tap, $100; labor to complete wiring, $120; tapping fee, $55; electrical parts, $85. Aside from these costs, actually seeing everything now installed is priceless.

Monday morning comes around. You’ve called the plant operator to ask if the speed switch was working. He said it was working just fine, but an hour later, he calls back to inform you that a rock rolled off the conveyor and broke the speed switch off the shaft. So, you are probably wondering what your alternatives are now, right?

Photo 7 shows a magnetically mounted speed switch. It’s very simple in design with one unique advantage over the competition — it is attached by a very powerful neodymium magnet that simply snaps onto the shaft you need to monitor. This product truly can be called an “out-of the-box” installation.

Newer isn’t always better

This has nothing to do with the high-tech automation we currently have installed in our plants, but it is worth noting. What we have here (refer to Photo 8) is the 100-year-old, spring-wound timer. It is used for fans, lights, and just about anything you need to control. And you know the best part about it? The fact that there are no chips, resistors, capacitors, filters, diodes, rectifiers, etc. — nothing to burn out or reprogram. You do not need a neutral for it to work. It’s this simple — just put it in, connect the wires, and use it. You can buy them scaled in seconds, minutes, and hours.

I sometimes compare items such as these to my wife. There is nothing on the market as reliable, straightforward, low-maintenance, trustworthy, etc. But just as I sometimes do with my wife, I sometimes take such things for granted. Go ahead — buy the more expensive, more exotic piece of equipment. You will be back someday, and she will be waiting on you at half of what you paid for the Ms. Everything Switch. Look at her. Isn’t she beautiful now? Just wait until later when the switch is too complex to figure out how to fix it. This is why you need to make sure you only get what you need with automation. If you need some high-tech features, get them. But getting much more than you need can only lead to trouble.

Leveraging the right amount of automation

Our industry is currently undergoing various stages of plant automation. Some operations are crawling with it, others have learned to walk — and still others are running ahead with it. As a whole, the cost of most equipment used in automation has never been more attractive. We have a tremendous opportunity to justify plant automation, no matter what size the plant is. Just think back a few years ago when the major hurdle was to convince ourselves that the cost of automation was justifiable. Today, whether we are crushing one onion or many onions, automation is still a small price to pay compared to the increased production.

The one thing we need to do is to concentrate on installation and planning. Use good judgment to determine if certain equipment is really worth the cost. Choose sensors that are easy to install and easy to replace. Select equipment that is “standalone” so that we don’t pollute our systems with add-ons. When determining options on drive components, don’t go overboard. Remember, keep it simple. Automation will survive as long as we instill confidence in the people who install the automation and the people who use it.

We have the control equipment available today to accomplish tasks that only a few years ago were either impossible or much too costly. What about tomorrow? I am sure that during the years ahead, our plants will be monitored by smart cameras that can perform the same tasks that our own eyes do. Reliable wireless instruments will most likely replace the wiring we perform today. The technology will increase to solve just about every possible problem that we currently have. But for now, we have to use what is available.

As my career has been influenced by automation, I have been inspired by the words of two men. The first is George Bernard Shaw, who wrote, “Some men see things as they are and ask, ‘Why?’ I dream things that never were and ask why not?” The second is Paul Detwiler Jr., chairman of the board, New Enterprise Stone and Lime Co., who says, “If you’re going to run a railroad, you’re going to have wrecks.”

This is true with automation sometimes. Simply put, the success of automation will be enhanced when the installation is as uncomplicated as possible.

Further, when the wrecks happen, this success is amplified when the train is back on the track swiftly. Our minds have a huge capacity for innovative thinking. But the only time we can exercise our thoughts is when nothing is detracting our attention. If we are bogged down with maintaining an overcomplicated system, we have little time to dream of the future plans.

Gary Wineland is the director of plant maintenance for New Enterprise Stone & Lime Co., Inc., where he previously served as director of automation. He is also the president of Applied Innovative Solutions, LLC, a manufacturer of products used in plant automation.

Reprinted from Aggregates Manager Magazine
November 2005