My service panel is over a concrete patio, it was improperly grounded in the crawlspace. I cannot drive a proper ground rod in the crawlspace so I would like to drive one into the earth on the side of the house. This would place the rod about 20' away.

Is there a limit to the length of the ground conductor from the panel to the rod?

  • I don't think that there is a limit of how far away the rod can be, that I know of. That being said, you will want to make sure you also don't put it too close to some other lines, like your gas line or cable line, and make sure you don't dig into a sprinkler system either if you have one. I don't know that it's a huge deal on how close you are to those lines but you probably don't want to find out. It is also usually recommended to have two ground rods, the minimum distance those need to be apart from each other is 6 ft. The main concern with putting in a ground rod is the connection. You w – chh Jun 28 '13 at 19:26
  • 1
    Have you considered drilling a small hole in your concrete patio and driving the ground rod there? – Philip Ngai Jul 29 '13 at 16:22

This answer is mostly based on the United States electrical system, and the answer may vary depending on where you are.

The NEC code specifies that a solid copper wire used to connect to a ground rod must be at least either #6 or #8 gauge (depending on the size of your electrical service cable). #6 cable cable will always satisfy the sizing requirement, though in some cases larger is desirable. The NEC doesn't specify a limit as to the maximum length. Of course, shorter is better.

Aluminum cable is permissible with larger minimum sizes, though I'd avoid it due its corrosion properties.

The grounding system serves a few different purposes:

  1. Make the voltage of the land around your house be at approximately the same voltage as your power line neutral. This means that you can't shock yourself by holding an appliance in one hand and touching the earth with your feet.
  2. A fault path during lightning storms. When lightning strikes near your house, it energizes your electrical lines. Surge suppressors will try to shunt the voltage to the neutral or ground line. Having a good ground will maximize the power transfer out of your house.
  3. Safety. If the neutral going to your house is disconnected and you don't have a ground to neutral bond in your main panel, both prongs of each electric outlet in your house will have the full line voltage on it. Older appliances will connect their chassis to neutral, causing their chassis to hold 120V. With the grounding system properly connected, the ground will pull the neutral closer to 0 V, reducing the risk of shock.
  4. Radio antenna. Many radio antennas use the ground to help transmit their signals.
  5. Noise reduction. The power lines have a lot of "noise" on them. Having a good ground connection will reduce the noise that your equipment has to deal with.

So, what resistance to ground should you want? The smaller, the better. The electrical code states that with one ground rod, it must have a maximum resistance of 25 ohms to the earth. According to a Fluke brochure, you should try to have a ground to earth resistance of less than 25 ohms, or less than 5 ohms for sensitive electrical equipment.

#6 gauge solid copper is approx 0.4 ohm/thousand feet, so having a run of thirty feet will add minimal extra resistance (about 0.008 ohm). But, it will somewhat reduce the effectiveness of your system during a lightning storm. For the best lightning protection, your grounding wires should not have any sharp bends. This is because lightning is a very high frequency signal, and the wire's impedance increases with frequency. The 0.4 ohm/thousand feet figure is only valid at DC (zero frequency). While sharp bends do not increase the DC resistance, it does increase the high frequency impedance.

As far as a suggestion, for a "normal" house, I'd use solid #4 copper to two 8 foot long copper-plated ground rods which are placed 16 feet apart, and driven so that they are completely underground. Welding your ground wire to your ground rods is best (for example, with a product such as CadWeld's One Shot), but "Acorn" clamps are acceptable to use. The welded connection is more corrosion resistant than the mechanical clamp conection. Also, ensure that your water pipes are also attached to your grounding system. Your neutral should connect to the grounding system at the service entrance (generally the main breaker panel), and in no other location.

There is a limit on the length of the wire connecting a cable TV co-ax shield to your house's grounding system. It must be no more than 20 ft long, unless there is an additional ground rod (see NEC 220.100 for details), though I'm not sure if this would be the distance to your main panel, or to your ground rod.

There are many other rules that I have not mentioned (read the NEC book (NFPA 70) or your local code for details). As always, use caution when working around electrical systems.

  • It is also good to use a welded bond between the ground conductor and the ground rod - something like the Cadweld One Shot – Eric Gunnerson Jul 29 '13 at 4:20
  • NEC 2014 250.66(A) says the ground conductor doesn't need to be larger than a #6, which contradicts your "#6 gauge, or larger" in the second paragraph. – Dave Sep 13 '15 at 18:03
  • @Dave, I agree that it isn't ever required to be larger than #6 copper, but it's OK for it to be larger. Also, #4 isn't that much more expensive (since the length should be short), will last longer and will provide a better ground. Looking at the code again, I see that there are some conditions where #8 is acceptable which I did not mention. Also, aluminum is permissible, but probably a bad idea due to corrosion issues. – Pigrew Sep 13 '15 at 20:51
  • @Pigrew could you explain why lightning is a high frequency signal. – Kris Sep 14 '15 at 2:36
  • 1
    @Kris: First, understand that normally when we say frequency, we are talking about sine waves. But, lightning isn't a sine wave. It abruptly turns on, and then abruptly turns off. In order to figure out the frequency, we must mathematically represent the lightning strike as a sum of sine waves at various frequencies. The abruptness of the strike causes there to be frequency components above what you would expect (1 divided by the strike duration). The duration may be 30 us (equiv to 30kHz), but there are components at much higher frequencies (>100 kHz), limiting skin depth to <<1mm. – Pigrew Sep 14 '15 at 17:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.