787 Fuel Burn Performance & Public Data Limitations

There is much good to be discovered in the US DoT’s Form 41 data. But it is important to understand the limits of this source. As an example we are going to take a look at the data this source has on the 787. Recently FlightGlobal published (Nov 18-24) an excellent article on a series of interviews they undertook with airlines operating the 787. The overall feedback from airlines is positive. No doubt this gives Boeing a lot of comfort. Airlines report the 787 is a fuel efficient aircraft – with 20% savings over what the 767-300ER achieved apparently quite common. This is the number Boeing promised their customers. However when one looks at data from the DoT Form 41 there is a very different story to be told. Take a look at the table below. The DoT only has data being reported by United Airlines, since it is the sole US-based 787 operator. Fortunately United flies the 767-300ER also, allowing for an even comparison. The columns in blue are the actual data from Form 41. Definitions of each column are shown below the table. In order to try to even the comparison as much as possible we calculated the fuel use by fight hour and then by seat. Clearly the 787, according to this data, is nowhere near 20% more fuel efficient than the 767. United claims they were not contacted to be part of FlightGlobal's 787 story. So now what? Did airlines lie to FlightGlobal? We sent our table to Boeing and United, asking for help understanding what this might mean. Boeing provided the following response: “On average, 787 customers are seeing a 20 percent fuel advantage over their similarly-sized airplanes without efficiency upgrades. The variance in percentages is due to several factors including route structure and engine type. The longer the flight, the better the fuel efficiency. For example, a shorter-range mission won’t see much benefit, but a 6,000nm flight can grow to more than 20 percent. We won’t go into detail on specific customers, although ANA has been public about their greater than 20 percent fuel advantage.” As of this writing United has not responded. Meanwhile, given the obvious discrepancy in the data we also turned to masFlight, the leading source of airline operational data. They also took a look at our chart and then went off to do some digging. This is what masFlight came up with. The table shows data for a twelve month period on three routes using aircraft performance data, allowing for a fair competitive comparison carrying the same payload. The table shows that, in fact, the 787 is regularly producing a better fuel burn than the 767-300ER. masFlight estimates that, on average, for the routes shown in the table, the 787 operates with a fuel burn improvement of 15.3%. Given the stage lengths and taking Boeing’s statement about 6,000nm above, the numbers are plausible. This is not a 20% saving as stated by airlines in the FlightGlobal story. Though the differences are much smaller than we see from Form 41; 5% is still significant. How to explain this difference? As with all data sources there are variances that are not obvious. For the Form 41 a reasonable suggestion might be that the data filed by United does not allow for a fair and equal comparison when we group all the data. The Form 41 remains a useful tool for all sorts of analysis. But one must understand that it does have limits. What we need to do there is ensure the comparisons are done on a per route basis, normalizing payload or Zero Fuel Weight and air miles flown to ensure comparable mission profile for both aircraft. Indeed this is what we accomplish with the masFlight data. How then to explain the difference between masFlight data and what airlines told FlightGlobal? We believe the differences can be explained by, again, understanding data sources and inputs. For an airline, like LAN, flying relatively new 767-300ERs the improvements from the 787 are going to be smaller than 20%. On United’s routes, we may be comparing older 767-300ERs with new 787s where fuel burn savings will be higher. In addition if the 767-300ER has winglets, then its fuel burn delta with the 787 will be less than 20%. Nuances are not insignificant; rather they can add up to show big differences. When Boeing first spoke about the 20% savings, no winglets were being considered on the 767, so the number was a reasonable target. In conclusion what we learn from this is that when comparisons need to be made it is important to ensure that data sources are as clean as possible. Fair comparisons require exactly even playing fields. This is an industry that does not have a shortage of data. But when data is pulled for analysis, if it doesn’t pass the initial smell test, it’s not necessarily wrong. It most likely means that one has to seek out another view and try again.

No results