This data is all based on calculations, no test plots planted, no actual measurement of water applied to crops(rain or irrigation), no record of NPK applied, no actual measurement of liters of oil produced. Thus no direct correlation between water applied and oil out. The authors obtained rainfall data and evaporation data, then suveyed crop yields to get a water to oil yield ratio (they call it water footprint). This method of analysis has a large margin for error.

Remember that studies do not give results, only point a possible direction that real empirical work (testing) may produce. This study does a lot of assuming according to the "supporting information" shown on the right sidebar of the article. Too much assuming means the conclusion is not likely to be accurate.

yes, good point. This study in Egypt tried growing them under water stress and used between 300 and 400 L / per kg of oil.

Response of Jatropha curcas L. to water deficit: Yield, water use efficiency and oilseed characteristics
Abdrabbo A. Abou Kheiraa, Nahed M.M. Attab
Water Management Research Institute, National Water Research Center, Delta Barrage, P.O. Box 13621/5, Egypt
Oil and Fat Research Department, Food Technology Research Institute, Agricultural Research Center, Giza, Egypt
ABSTRACT
Field experiment was carried out at Enshas Experiment Station; Jatropha was transplanted
and treated after the second year of the transplanting by different amounts of water stress,
viz. 125%, 100%, 75% and 50% of potential evapotranspiration (ETp). The study aims to
ensure the multiple benefits of Jatropha and its suitability under Egypt’s climate in unused
lands under scarce water conditions. The results revealed that the average water
consumption rate of Jatropha is 6 L per week throughout the growing season, which means
that Jatropha can survive and produce full yield with high quality seeds under minimum
water requirements compared to other crops. The yield of extracted oil was 85.5, 175.2, 90.5
and 66.5 kg/ha at 125%, 100%, 75% and 50% of ETp, respectively. The lowest values of total
lipid (oil) (25% and 24.5% of Jatropha seeds) were recorded with Jatropha trees that were
irrigated by 125% and 50% of ETp, respectively. On the other hand, the treatment that was
irrigated by 100% of ETp (control) recorded the highest value of total oil in the seeds
(29.93%). The results also revealed that there are no significant differences among the
values of the determined oil characteristics due to different water stress ratios. From the
results, it could be concluded that the highest characteristics of Jatropha seed oil were
recorded with 100% of ETp. In addition water stress had no significant effect on the fatty
acid composition of Jatropha seed oil.

Jon R. Luoma at Yale e360 has a excellent report on the problems with jatropha and the companies that are trying to make it work. There is a big conflict brewing in India where companies are buying off gov't officials to take grazelands from villagers to grow biofuels for export.

Jon R. Luoma at Yale e360 has a excellent report on the problems with jatropha and the companies that are trying to make it work. There is a big conflict brewing in India where companies are buying off gov't officials to take grazelands from villagers to grow biofuels for export.