Once a year BP plc publishes their Statistical Review of World Energy, as they have done for 65 years, now. Recent editions have been digital and anodyne, with lots of mini-analyses and charts and positive messages about the petroleum industry. Whenever energy researchers ask questions, they are invariably directed to take a look at the BP report, as it is considered trustworthy and sound. Good scientists always try to find alternative sources of data, but it can be hard comparing the BP Stat Rev with other numerical offerings, partly because of the general lack of drill-down in-depth figures. Two other reputable data sources are the US Energy Information Administration (EIA) and the JODI Oil initiative. I have already looked at EIA data and data from the National Energy Board (NEB) of Canada recently in order to check on the risks of Peak Oil. Now I’m diving into JODI.
Two of my concerns of the week are to try to understand the status and health of the global economy – which can be seen through the lens of overall consumption of hydrocarbons; and to see if there are changes happening in relative demand levels for the different kinds of hydrocarbons – as this could indicate a transition towards a lower carbon economy. The BP Stat Rev of June 2016 offers an interesting table on Page 13 – “Oil: Regional consumption – by product group”, which breaks down hydrocarbon demand into four main categories : Light distillates, Middle distillates, Fuel oil and Other. The “Other” category for BP includes LPG – Liquefied Petroleum Gases, a blend of mostly propane and butanes (carbon chain C3 and C4), which are gaseous and not liquid at normal room temperature and pressure – so strictly speaking aren’t actually oil. They also have different sources from various process units within petroleum refinery and Natural Gas processing plants. The “Other” category also includes refinery gas – mostly methane and ethane (carbon chain C1 and C2), and hydrogen (H2); and presumably fuel additives and improvers made from otherwise unwanted gubbins at the petrorefinery.
Not by coincidence, the JODI Oil database, in its Secondary data table, also offers a breakdown of hydrocarbon demand from refinery into categories almost analagous to the BP groupings – LPG, Gasoline, Naphtha, Kerosenes, Gas/Diesel oil, Fuel oil, and Other products; where LPG added to Other should be the same as BP’s “Other” category, Gasoline added to Naphtha should be equivalent to BP’s “Light distillates”; and Kerosenes added to Gas/Diesel oil should be analagous to BP’s “Middle distillates. So I set out to average the JODI Oil data, day-weighting the monthly data records, to see if I could replicate the BP Stat Rev Page 13.
Very few of the data points matched BP’s report. I suspect this is partly due to averaging issues – I expect BP has access to daily demand figures, (although I can’t be sure, and I don’t know their data sources); whereas the JODI Oil data is presented as monthly averages for daily demand. However, there are a lot of figures in the BP report that are high compared to the JODI Oil database. This can only partly be due to the fact that not all countries are reporting to JODI – four countries in the Commonwealth of Indepdendent States (CIS) – formerly known as “Former Soviet Union” – are not reporting, for example. I’m wondering if this over-reporting in the BP report might be due to differences in the way that stock transfers are handled – perhaps demand for refinery products that are intended for storage purposes rather than direct consumption is included in the BP data, but not in JODI – but at the moment I don’t have any relevant information with which to confirm or deny this concept.
Anyway, the data is very close between BP and JODI for the United States in recent years, and there are some other lines where there is some agreement (for example – Fuel oil in Japan, and Light distillates in China), so I am going to take this as an indication that I understand the JODI Oil data sufficiently well to be able to look at monthly refinery demand, refinery output and oil production for each region and hopefully reach some useful conclusions.