Empirical Analysis of the Multi-Mission Radioisotope Thermoelectric Generator Qualification Unit Operated at a Low Thermal Inventory with Potential for Improved End-of-Life Power

Performance predictions for the first multi-mission radioisotope thermoelectric generator (MMRTG) flight unit and engineering unit were recently reported. Both units were produced and operated/tested within specifications [i.e., nominal thermal inventory  = 2000 W(thermal)]. In an attempt to study the effect of a deep space cruise on an MMRTG that has been operational for 6.25 years (2.25 years storage + 4 years cruise), the qualification unit (QU) was placed on life test with a below-specification thermal inventory of 1904 W(thermal). Analysis indicates that loading an MMRTG with a lower thermal inventory may result in less power at the beginning-of-life but more power at the end-of-design-life (EODL). The lower thermal inventory in the QU produces a lower operating temperature, which appears to cause a significant reduction in the degradation rate of the thermoelectric couples. Preliminary calculations indicate that a thermal inventory of 1904 W(thermal) could result in a 9 W(electric) power boost at EODL [i.e., 84 W(electric)], which is a 12% improvement over the first MMRTG flight unit and engineering unit predictions. Preliminary degradation analysis suggests that a 1904 W(thermal) unit will begin to produce more power than a 2027 W(thermal) unit approximately 4 years after fueling. This suggests that missions with a primary power requirement more than 4 years after fueling would benefit from a lower thermal inventory. In addition, using a lower thermal inventory has significant benefits for ²³⁸Pu stockpile management and may allow for additional MMRTGs to be fueled from our current reserves. Conclusions and hypotheses presented here should be considered preliminary because the QU data set is very small and there are some uncertainties regarding how early-life QU data will translate into later-life performance. More QU testing at a thermal inventory of 1904 W(thermal) is needed to prove that the preliminary conclusions presented here are valid.