Key to this blind spot is that the source material for most of the assessments is provided by a small number of data compilers, many if not all of them working in collaboration with the International Energy Agency (IEA). The data is voluntarily submitted by the industry in response to academic surveys. The nature and profile of the respondents is never publicly revealed, so that there is the potential for conflicts of interest to develop.

A further puzzle is how that data feeds into an organization called Ecoinvent, a Swiss-based non-profit founded in 1998 that dubs itself “the world’s most consistent and transparent life cycle inventory database”. This data is relied on by institutions worldwide, including the IPCC and IEA itself, to calculate their carbon footprint projections, including the sixth assessment report published as recently as March 2023.

Based on such data, the IPCC claims solar PV is 48 gCO2/kWh. But, as we’ll see below, a new investigation started by Italian researcher, Enrico Mariutti, suggests that the number is closer to between 170 and 250 gCO2/kWh, depending on the energy mix used to power PV production. If this estimate is accurate, solar would not compare favorably with natural gas, which is around 50 gCO2/kWh with carbon capture, and 400 to 500 without.

Over the course of a four-month investigation, Environmental Progress has confirmed that Ecoinvent — perhaps the world’s largest database on the environmental impact of renewables — has no data from China about its photovoltaic industry. Meanwhile, the ultimate source of the IEA’s supposedly public data on PV carbon intensity is confidential and the data, therefore, unverifiable.