Comparing IR camera resolutions - what high resolution buys you

By Dr. Robert P. Madding, Director
Infrared Training Center

Why High Resolution? A higher resolution camera means you will find smaller problems at greater distances. You can find significant problems that could be missed with a lower resolution camera.

IR survey efficiency is greatly improved

You can cover same surface area with ¼ to 1/16 the number of images. This means faster surveys and faster report generation.

       
Infrared pixel counts for Low, Medium, and High resolution infrared cameras (click image for larger version).

The Hi Res can find the needle in the haystack while looking at the entire haystack.

The Hi Res camera can be 4 times the distance of a Low Res and 2 times the distance of a Mid Res system and get the same image detail. At 1/2X distance, it takes 4 pictures to cover the same area; at 1/4X distance, it takes 16 pictures to cover the same area; think of the savings in capture and reporting time; and image interpretation just got easier.

Small, distant targets

The Hi Res camera really comes through for small, distant targets. Substations, switchyard, transmission and distribution lines are littered with this type of target. One could argue you could use a medium res camera with a 2X lens; but you now have ½ the FOV so you have to scan more to cover the same area—and that takes longer.

Examples

Here are a few examples comparing the different resolutions and thermal measurements.

Utility pole for 23 KV distribution

Here is a typical utility pole for 23 KV distribution. It has three switch disconnects. One of them has a problem. You’re trying to drive down the road at 20 mph (32 km/h) while your thermographer is panning the lines for potential problems. You need to go slow enough to find them, yet fast enough to finish as efficiently as possible. Normally, the utility does this very early in the morning starting about 0430 and finishing before traffic gets too heavy. The better your resolution with a wide field of view, the faster you can go without missing anything. Telescopes will work here but now you have to slow down even more.

An example, a switch disconnect at 12 yards.

It takes a Hi Res camera to see the bolted plate connection is problematic. Cycling the switch won’t do any good. You have to get in the bucket and repair the connection.

     
Low, Med, and High res IR camera images at 12 yards.

The image below is a magnified detail between the medium and high res camera images along with a table describing the temperature measurement differences.

Magnified view and temperature table for switch disconnect.

We used the same emissivity, 0.95, assuredly higher than the actual emissivity. Load correction software gives about 3X delta-T when load is doubled. This is more nearly correct than the very erroneous “square law” which would give 4X, 33% higher. The so-called square law presumes that delta-T is linear with heat loss and power dissipated varies as the square of the current (load). The latter is true, but the former is quite untrue as heat transfer by radiation varies as T⁴. Ignoring radiant heat transfer means our IR camera would not be able see the problem!

The key factor is when measuring at moderate loads, the uncertainty is greatly magnified when simulating higher load conditions, or actually going to a higher load.

115 KV to 23 KV transformer

This transformer supplies power to houses and light commercial industry. An unscheduled outage could cause customers their power for several hours. The grid does a good job of re-routing, but your lights will go off for a while when this baby goes. And the tank below the wires is full of mineral oil, very flammable.

   
Visual image of substation transformer low side and thermal images showing "barber poling"

This is the low side of a 115 KV to 23 KV transformer under 40% load as measured by National Grid. The Low Res and Med Res show the problem on two phases, but they don’t have the resolution to get good temperatures. And the Hi Res is the only camera to indicate this problem is beginning on the third phase as highlighted by the oval.

It is difficult to tell with the Low Res at this range of 11 yards what is going on. Both the Med Res and Hi Res clearly show “stranding” or barber poling where some of the strands of the cable are disconnected or very poorly connected and the remaining ones are overloaded and overheated. The Hi Res gives more detail - one can almost make out individual strands. This is a very serious problem and should be repaired asap. The cable is about ¾ inch diameter at 11 yards.

Temperature Data “Stranded” Cable at 11 Yards

The temperature rise was 211 F (117 Celsius). Considering we used 0.95 emissivity certainly higher than the actual emissivity and that the unit was under 40% load, this is a definite critical problem. Load correction shows the temperature approaching 1000 F (500 Celsius) for full load conditions at the same emissivity.

PMH 23 KV connections

This is one of those green boxes on the ground that we occasionally walk by without giving them a second look. Some contain transformers, others contain buried connections such as this one.

Qualified personnel with proper PPE can approach within 2.5 feet ( 75 cm) of this PMH. Thermographers not so qualified can approach within 10 feet (3, 5 meters) under supervision. Frankly, the farther one can be from live switchgear, the better in terms of safety. Arc flash explosion effects drop as the square of the distance. We help minimize risk by staying back and keeping the cover off as little as possible. A high resolution camera really helps here. We also noticed the Med Res should give the same reading as the Hi Res as the target is plenty large enough. But the hot spot doesn’t always match target size. It can be much smaller. So, its not always small targets at a distance that need good resolution, but larger target closer in with small hot spots can also benefit from a higher resolution camera.

PMH 23 KV connections

It is very important here to try and diagnose where the problem exists. At the top is a bolted plate connection. At the bottom is a compression connection to the underground cable. If the compression connection is the problem, and it looks like it is, repairmen must dig up cable. If it is the bolted plate connection, it can be repaired without digging.

   
PMH 23 KV temperatures

Thermal measurements  PMH 23 KV connection at 10 feet – 3 meters

Though the temperatures vary a small amount, when calculating for full load from 50% load, we multiply delta-T by about 3. Errors in measurement are magnified.

Summary

• Higher resolution means better ability to spot small targets at larger distances
  Find problems you would otherwise miss completely, improve diagnoses with better image detail

• Higher resolution means better measurement accuracy of small targets

• Higher resolution means larger field of view than ever before for spotting small targets
  Take fewer images. Improve your survey and reporting efficiency with Hi Res

The cameras used for the imaging and thermal measurements in this report may be found in the following table.

In Appreciation

Thanks, Ken Leonard and Donald Smith of Progress Energy for going above and beyond the call of duty to help us get the pole and PMH images. Ken and his co-thermographer, Donald, spent three days prior to our arrival to scope out possible problem targets. Then they spent two more days with us and a video crew shooting images.

Thanks, Carmine Luongo and Mark Moriarty of National Grid for allowing us to enter a couple of substations on very short notice. We found the barber pole connections on the transformer, a critical problem, and sent a report to Mark right away. Now the lights will stay on in Lowell, MA.