Special Note: This phenomenon can occur in your plant, regardless of voltage or type of grounding.

A recent trade publication story blared the startling news: worker killed by machine, cause unknown. No one came right out and said that the machine started by itself. However, there were no witnesses, and an operator does not put half his body inside a machine and then reach around and press the start button. And even if he did, the door switch would prevent the machine from starting. So what happened?

It is a fact that machinery can start by itself. We are not talking about operator error, someone committing a foolish act and then denying it, or a lockout/tag out violation. We are talking about a machine sitting idle and then starting to run without any intervention by an operator. I personally have been associated with two such occurrences. Moreover, the situation is so prevalent that some machine manufacturers put signage in their machinery control cabinets with instructions attempting to reduce this phenomenon.

Because many machines require operators to stick part of their bodies inside the machinery, usually to unload, this fact should scare you. The phenomenon doesn’t happen every day, but it happens more often than one would imagine. Proven frequency should take a back seat to the simple fact that it does occur. When it does happen it is often misdiagnosed. After all, when people think they see something that they find hard to believe themselves, they don’t always share the experience with others, or they simply don’t believe it. And those who know why it happens find it difficult to write a technical explanation that management will read and understand.

Basics: the ground

The problem of any machine starting up by itself is most often caused by two electrical grounds. These grounds don’t normally happen together. One ground is either presented by the utility company or is an undetected fault. When the second ground occurs, the machine starts.

When we speak of grounds or grounding something, we mean that we are connecting something to earth ground. In other words, we physically connect the item in question to the earth through some kind of electrical conductor. This might mean that we drive a rod into the ground and make an electrical connection to it, or we might use something already in the ground such as water piping.

Why do we need to do this? There are two primary reasons. One purpose is to protect people that use electrical equipment. The earth is a good conductor of electricity, with a large capacity to store electrons and plenty of moisture and impurities. Place a difference of potential across it (voltage) and it will conduct. So, if you are in contact with (the) ground, which you usually are, and come into contact with an electrical wire that is energized, electrical current will pass through you to (the) ground and may cause you harm. If, however, someone has already grounded the electrical device that you touched, then the potential of the device and ground are the same. With no difference of potential, no current will flow.

These are intentional, or safety grounds. Safety grounds are not connected to energized circuits. They are connected to the cases of electrical equipment or conduits for electrical wires so that if there is ever an unwanted path of current to the case or conduit (fault grounds), you will not receive a shock if you touch it (i.e., the current will not flow through you).

The second type of ground actually connects an electrical circuit to the earth. This may be intentional or due to a fault, and this is the kind of grounding that causes our problem.

There are typically three ways to create this dangerous type of grounding. Two come from the utility company.

1. For reasons that are unclear, some utility companies at one time had the practice of intentionally grounding one phase of the power they provided your plant. They only did this on voltages in the 230-volt range, called corner grounding, and they no longer follow this practice. But don’t bother asking them to change those older installations—they generally will not. This may be because to change it might acknowledge that it was wrong in the first place, opening up litigation for past accidents. Memphis, TN; St. Louis, Louisville, KY; and possibly others currently provide grounded corner electrical service to industrial plants.
2. Another more common type of this grounding provided by many utility companies is what is called a stinger, or high leg. The utility company, providing 230 volts through a three-phase transformer assembly, grounds the center tap of one phase of the transformer. This sets up a cheap way to provide both 230 volts and 115-volt lighting power through the same transformer bank. By connecting loads to either of the two legs that are connected to the grounded phase, and the other side connected to earth ground, 115 volts is provided. By connecting loads between any two phases, 230 volts is provided. Unfortunately, when you read these three phases to ground, two read 115, but the third reads 190 volts. This third leg, the stinger, as with the grounded leg mentioned in paragraph 1 above, may cause serious problems.
3. The third way to establish this type of grounding is through what is called a fault ground. A fault ground is an unintentional, unwanted path of current to ground. This can be caused by deteriorated insulation, moisture, or wear. It has the exact same effect of those grounds mentioned above, except that the ground occurs accidentally. If you wash down your machinery and water gets into a connection box, the wiring is old and eventually it or the insulation cracks and falls off, or chafing occurs due to machine vibration wearing insulation away, a path of current flow to ground is created.

To understand how this becomes a problem, it is necessary to understand that for electrons to flow, they must have a complete path from one side of the source back to the other side of the source. In other words, one ground does not create a problem, just the potential for a problem. There must be a path from the power source to ground, and then back from ground to the other side of the power source.

Bypassing “start”

So let’s look at several hypothetical situations. You operate a plant that is provided either a corner ground or stinger leg from the utility company. Or you operate a plant that is not grounded from the city, but a fault ground has developed somewhere in your electrical system (this can happen with electrical systems of any voltage). In any of these scenarios, when a second ground occurs due to moisture, wear, deterioration, etc., the machine will immediately start if that second ground occurs in certain locations in the control circuitry.
Figure 1


Figures 1 and 2 show what is known as a schematic wiring diagram. It is called this because it is used to show the scheme of a circuit, i.e., what happens when a button is pushed, a contact opens, a fuse blows, etc. The circuit drawn is a simplified drawing of a motor starter. Every motor in your plant that drives major machinery is controlled by a circuit similar to this.

In both figures, the points labeled L1 and L2 represent the line leads that are connected to your plant electrical system, usually through a circuit breaker (switch). The boxes labeled F1 and F2 represent fuses, each of which is simply a wire that will conduct current unless it gets hot and melts, which happens if too much electrical current is present. The symbol with a line through it is a contact, such as a vibration switch. There is a stop pushbutton that lets current pass unless it is pushed, and a start button that will not let current pass unless it is pushed. To the right of the start button is the symbol for a relay coil. I labeled it CR for Control Relay. When current passes through this coil it becomes an electro-magnet, magnetically pulling in a device that operates some contacts. These contacts will be labeled the same as the coil. One of those contacts is shown in parallel with the start switch. When it closes, the operator can quit pushing the start switch and the coil will stay energized because the CR contact is providing a path of current around the start switch.

So when the start button is pressed, a complete circuit is made from L1 through the CR coil to L2. The CR contact bypassing the start switch is closed when the coil picks up so that the start button can be released without de-energizing the coil. Some other contacts not shown on this drawing for clarity will also close. These are in the motor circuit that we are controlling. When these close the motor starts. The motor then runs until someone pushes the stop button, a fuse blows, the vibration switch opens, etc. When this occurs, the coil becomes de-energized (no longer a magnet) dropping out its contacts. The contact around the start switch and the motor contacts open. The motor stops.

Now let’s look at this drawing again, only let’s assume that L1 is connected to ground. It doesn’t matter whether it is grounded by the city as in a corner ground, or a fault ground has occurred in your plant. There is a path of current from L1 to ground. Remember, there must be a complete path of current from L1 to L2 for current to flow. With only one ground, we have a path from L1 to ground, but do not yet have a way for that current to get from ground to L2.
Now, let’s assume that a second ground (fault) occurs at the arrow in Figure 2, to the right of the start switch.
Figure 2

Due to moisture, a bare wire, etc., the circuitry just to the right of the start switch is connected to ground as well. Essentially, the right side of the start switch (L2) is now connected through ground to L1. Put another way, I now have a path of current from L1 through ground into the wire at the arrow (fault ground), through my coil and on to L2. The start switch is bypassed altogether. The machine starts, seemingly by itself. Scary isn’t it?

Hazard remediation

While it is a little more complicated to describe, a very similar thing occurs if you have a stinger or high leg providing power to the control circuit.

So how do you know if you have either a grounded line lead or a stinger leg? If each of the three phases of power coming into the plant is read (tested) with a voltmeter to ground, a ground or stinger leg is easily detected. In the case of a stinger or high leg, two phases will read approximately 115 volts to ground, and the stinger will read approximately 190 volts. In the case of a grounded leg, two legs will read full line voltage (230, 480, etc.) and the other leg will read 0. The 0 reading is on the grounded phase because there is no difference of potential between a leg that is grounded and ground itself. If no ground or stinger is present, each leg read to ground will indicate approximately half of full-line voltage.

If you perform the test above, and determine a ground or stinger leg is present, eliminating this potential hazard from your plant is not too difficult. First, if the city is providing either a grounded corner or a stinger leg to your plant, it is vital that this grounded or stinger leg not be connected to any motor control circuits. In three-phase motors, it takes all three phases to drive the motor properly, but it only takes two to energize the control circuit. Just make sure the other two are used for the control circuit. Second, if your ground is caused by a fault in the plant, it must be located and eliminated. Any maintenance engineer with good electrical skills should know how to perform this task.

Once every motor control circuit has been checked to ensure that it is not hooked to the stinger or grounded leg, on-going training is necessary. Otherwise, when a motor is replaced or a new piece of machinery is added, the potential to connect it to the wrong leg is ever present.

Additionally, it is necessary to routinely check for the presence of a fault ground (at least once a week). This only takes a minute, and it is an essential preventative step. Since a grounded phase shows up anywhere that phase is tested, it is only a matter of going to any three-phase controller or power panel and reading each phase to ground. If this isn’t done frequently, when a fault occurs and remains undetected, it is only a matter of time until that second fault occurs and creates real problems for the plant.

A 40-year industry veteran, Earl Winter is a former engineering vice president for a large textile rental company. He is currently president of Earl Winter Engineering Associates, Inc. Contact him at 770/402-7820 or e-mail earl@earlwinter.com.

Home • Boiler Program • Recruiting • Maintenance