A continuous output (non-pulsed) DD 2.5 MeV unit of up to 1x108 n/s. The choice of output should consider that the output and hence cost can be optimized to match the 252Cf output at the end of the useful decay time. This may be 2.6 years or earlier if half the original activity of the 252Cf source cannot be tolerated. So a 5x107 n/s specification for the neutron generator may be attractive as it halves the high voltage power requirement.
For this example we assume that the purchase price of the NSD neutron generator is twice that of the first load of 252Cf.
A non-pulsed DT 14 MeV NSD neuron generator is also included. At the same output specification the high voltage power supply is significantly reduced. This leads to a substantial cost reduction despite the costs associated with Tritium.
A wipe test every six months is usually mandated by radiation safety regulations for sealed isotopic sources. This cost includes the fee charged for the inspection and the cost when the inspection causes disruption to production.
Every 2.6 years the 252Cf must be topped up to compensate for the decay. Some applications may require more frequent top-ups in order to stay within an acceptable performance range.
The comment from a web page of a PGNAA system vendor: "At present the cost of a neutron generator is on the order of $100,000 with an expected life on the order of a couple years.” - This certainly does not apply to the NSD neutron generator.
In mid 2008 the US manufacturer of 252Cf (90% of the world supply) announced its intention to end production. The US Government has announced that it shall no longer fund 252Cf beyond 2009. While a consortium of private companies are working together to raise funding for continued production, it has already been expressed that the cost of 252Cf will increase at least a factor of 6 by 2011.
It is assumed in this example analysis that the Tritium gas mix will be recharged every 36 months. This is a rather conservative estimate. The interval may be 5 years. The recharge may be implemented as a return of the reaction chamber to the manufacturer or an exchange reaction chamber which can be quickly installed.
The degradation caused by accumulation of Helium as the reaction product accumulates will very gradually alter the performance. This can be compensated by use of reserve power. Together with very slow degradation of the reaction chamber we expect age effects to be significant by 50,000 hours of operation. Then an exchange reaction chamber can be installed and the original unit refurbished for further use.
Compare this NSD maintenance cycle with the replacement of a solid target sealed cartridge every 8000, 4000, 2000 hours or less. The old cartridge is destroyed in order to recover the Tritium. There really is no point in comparison but we have included a chart to make the point absolutely clear. The 15 year life cycle cost of Solid Target Neutron Generators is between 10 and 20 times more costly than that of an NSD Neutron Generator.
Only by purchasing and not using a sealed tube solid target neutron generator can one approach the economics of using an NSD neutron generator continuously.