Van de Graaff generator safety

Van de Graaff generator demonstrations can provide useful insights into electrical phenomena, which are at the same time memorable.

• It is essential the Van de Graaff generators for school science are obtained through reputable school science equipment suppliers. The sphere has a capacitance, and will store energy electrostatically as a result of the separation of charge. The energy stored electrostatically by the sphere should not exceed 0.5 J.
• Do not add devices to the sphere that increase the capacitance.
• Van de Graaff generators with mains powered pulleys must be electrically inspected and tested in the same way as other mains powered equipment.
• When carrying out the hair-standing-on-end demonstration, do it with one person at a time. After the demonstration, to avoid a sudden discharge, the person should take their hand off the sphere and touch the surface of a wood bench top (avoiding metal fittings such as gas taps). Alternatively, hand the charged person a wooden metre rule. After a few moments, they will be discharged.
• It is not advisable for people to participate in practical work with Van de Graaff generators if they have heart conditions, or a pacemaker, or other electronic medical equipment fitted.
• The electrical discharge from a Van de Graaff generator can wreck electronic circuits, so equipment such as computers and instrumentation with electronic circuits should be kept well away.

The Van de Graaff generators designed for schools are usually the triboelectric type; these are the most suitable. The transfer of charge is achieved by a rubber belt driven by a plastic pulley, with an arrangement of metal combs at either end of the belt. Charge is transferred to a metal sphere (a capacitor) and very high voltages are achieved between the sphere and ground, typically in the range 200 kV to 300 kV.

Using a Van de Graaff generator, one is quite likely to receive a short shock by accidental or intentional contact with the charged dome. An enquiry to CLEAPSS has revealed no recorded incident of direct injury caused by shocks from the correct use of school Van de Graaff generators. However, some people are more sensitive than others and can find the shocks very unpleasant and painful. For this reason, only volunteers should take direct part in the practical work.

The shock is a single unidirectional pulse of short duration; a capacitor discharge through the resistance of the body and contacts. Using rough values of sphere capacitance, 10 pF, and body resistance, 1000 ohms, the discharge time through a person will be of the order of nanoseconds and certainly less than a millisecond. The capacitance of the sphere needs to be low enough that at maximum voltage, the energy transferred by the discharge current through the body causes at most no more than an unpleasant or painful sensation. The current flowing and energy transferred to the body should be well below that which could cause any risk of ventricular fibrillation.

For an impulse current I Amps of short duration t seconds ( t < 10 ms) through the body, the principal factor for the initiation of ventricular fibrillation is the value of I  ×  t or I²  ×  t (IEC 2007). At high applied voltages, the resistance of the adult body (left hand to right hand) is at least 575 ohms for 95% of the population (IEC 2005). The total body resistance for children is expected to be higher but of the same order of magnitude. The IEC gives a threshold value of 'Specific Fibrillation Energy', for a 1 ms current impulse, of ValueExponent{2}{-3}A²s. Below this threshold there is no evidence of fibrillation. The Specific Fibrillation 'Energy' can be regarded as the energy dissipated per unit resistance of the body through which the current flows. Note that ‘specific’ here means ‘per unit resistance’ rather than ‘per unit mass’.

At 575 ohms and discharge time not exceeding 1 ms, the energy stored electrostatically by the capacitor would need to be at least 1.1 J to reach the Specific Fibrillation Energy. Although this is a very conservative estimate because the discharge time is likely to be much less than 1 ms, Van de Graaff generators that can store more than 1 J of energy electrostatically by the sphere should be avoided. A discharge of 1 J affects everybody severely (BSI 1991).

An estimate of the energy stored electrostatically by the sphere can be made by calculating the sphere capacitance, C = 4 π ε₀ r (where C is in farads, r is the sphere radius in metres and ε₀ is the permittivity of free space), estimating the voltage, V , using the length of spark gap, and calculating the energy E , in joules, from E = 0.5 CV².

Generally speaking, sphere diameters of Van de Graaff generators should be about 20 cm or less. Using data from one manufacturer’s specification, the sphere diameter is given as 20 cm and the maximum voltage 250 kV, the energy stored electrostatically would be 0.35 J, well below 1 J. If this is compared to a sphere diameter of, say, 25 cm and a maximum voltage of 350 kV, then energy stored electrostatically would be 0.85 J. This would still be below 1 J, but the shocks would be correspondingly more unpleasant and painful, and this may put off some people from using the generator.

If you wish to estimate the voltage across the sphere and ground, you can do this by finding the maximum spark length. Wait until the sphere is fully charged, then bring up a grounded sphere slowly until you obtain a spark discharge. This technique has limitations and you should do it carefully several times to find the maximum spark length. Do the test on a dry day with low relative humidity so the Van de Graaff generator is working at its best.

Note that the rule-of-thumb 3 kV/mm is only a reasonable rule for voltages below 100 kV.

References

IEC 2007. IEC/TS 60479-2:2007. Effects of human beings and livestock – Part 2: Special aspects.

IEC 2005. IEC/TS 60479-1:2005. Effects of human beings and livestock – Part 1: General aspects.

BSI 2002. BS EN 60052:2002. Voltage measurement by means of standard air gaps.

BSI 1991. BS5958-1:1991. Code of practice for control of undesirable static electricity Part 1: General considerations. [Replaced by PD CLC/TR 50404:2003 but remains current.]