A couple of years ago I made an unwise and possibly illegal connection to a power point (outlet), namely I supplied a resistive load (nominally a 250-watt heating element for a gas/electric fridge) via a series-connected diode, in order to halve the power dissipation of the load. You are probably wondering why! - it was to compensate for insufficient heat being delivered when I ran the fridge (designed for LPG) on natural gas.

A diagram of the culprit.

After a few weeks we had the most severe case of blue-green water you can imagine. When I realised that the unbalanced electrical circuit was to blame, I quickly removed the diode, and got the fridge converted to natural gas. Our house's neutral block is earthed to the copper cold water distribution pipes within my property.

There has been recent publicity in Australia about the problem of blue-green water. In the light of my dramatic experience, the following points should be noted:

1. Australian 240 volt ("low voltage", LV) electricity distribution systems are unusual by world standards, evolving from a historical need to distribute over large distances.

2. The very common three-phase method of LV delivery (with active and neutral single-phase wires to each customer), does have the potential to create locally raised earth potentials (AC?) if the loads on the three phases are not evenly matched.

3. I am told by an electrical engineer, Mr Alan Hutchinson of Plasmatronics, Fitzroy that such an imbalance can also inject a DC component if the waveform is not a pure sine wave.

4. Any switch-mode power supply that does not use zero-point switching has the potential to cause a DC imbalance if it is switching the positive and negative parts of the cycle at a slightly different phase angle. It will also distort the waveform.

5. Many households use water pipes to earth their household electrical wiring. The cold water pipes are usually used. Even if the hot pipes are used, they won't develop the blue-green problem, because the current can only flow along them lengthwise without ion exchange occurring. Surely ion-exchange can only occur where the pipe is buried in electrically-conductive soil. Most people are wise enough not to bury uninsulated hot water pipes in wet soil, so this probably explains why you apparently never get the problem in hot water pipes.

6. The increased prevalence of this problem in recent years seems to have paralleled the introduction of a wide variety of transformerless, offline switchmode power supplies, light dimmer switches, etc.

7. Stray currents are not the only cause, as it has been clearly demostrated that corrosive water alone can cause the blue-green problem. My argument is that electricity may be a bigger factor in a lot of cases than seems to be currently accepted.

8. The polarity of the diode may well influence the effect on the water: I believe the diagram accurately depicts the polarity that I used in my "experiment". I *think* the polarity as shown would make my cold water pipes electrically positive compared to the surrounding soil. The voltage induced is a result of the fact that the neutral wire back to the distribution transformer is not a perfect conductor: it has a resistance greater than zero.

---

Maybe householders with a blue-green water problem should, after checking for faulty appliances, consider disconnecting their house wiring neutral earthing point from the copper water pipes, and get a plumber and/or electrician to install a new electrical earth in damp ground as far from the copper water pipes as possible. This would have to be a lot cheaper than replacing the whole cold water supply with plastic pipes.