Assessing fuel burn inefficiencies in oceanic airspace

Stephen Builta, Purdue University


Increasing the efficiency of aircraft operations offers a shorter term solution to decreasing aircraft fuel burn than fleet replacement. By estimating the current airspace inefficiency, we can get an idea of the upper limit of savings. Oceanic airspace presents a unique opportunity for savings due to increased separation differences vs. overland flight. We assess fuel burn inefficiency by comparing estimated fuel burn for real world flights with the estimated optimal fuel burn. For computing fuel burn, we use the Base of Aircraft Data (BADA) with corrections based on research by Yoder (2005). Our fuel burn results show general agreement with Yoder’s results. Optimal operation depends on flying 4-D trajectories that use the least amount of fuel. We decompose optimal 4-D trajectories into vertical and horizontal components and analyze the inefficiencies of each separately. We use the concept of Specific Ground Range [Jensen, 2011], to find optimal altitudes and speeds. We combine the optimal altitudes and speeds with an aircraft proximity algorithm to find pairs of aircraft in a vertical blocking situations. To find the fuel optimal horizontal track in a wind field, we use methods from the field of Optimal Control. The original problem formulation can be transformed into a Two Point Boundary Value problem which we solve using MATLAB’s bvp4c function. From our set of flights, we hypothesized a scenario where aircraft stack in such a way that they cannot climb to their optimal altitudes because of separations standards. Using aircraft positions we find when aircraft were within separation standards and were blocked from climbing or descending to their optimal altitude. We split our inefficiency results into a blocked and non-blocked set to see if blocking had an effect on mean inefficiency. Our set of flights consisted of real world flights that flew through WATRS and CEP airspace regions during the month of April 2016. Using the optimal altitude for actual flight Mach profiles, we compute a mean inefficiency of 4.75% in WATRS and 4.50% in CEP, both of which are roughly 2 to 2.5 percentage points higher than studies using proprietary performance models and data. BADA overestimates optimal altitudes, leading to an overestimate in inefficiency. Inefficiency due to off-optimal speed for WATRS is 2.18% vs. 1.86% in CEP. Blocking events result in a 2.59 percentage point increase in mean inefficiency due to off-optimal altitude in WATRS flights, and a 1.21 percentage point increase in mean inefficiency due to off-optimal altitude in CEP flights. Using wind-optimal horizontal tracks gave a 1.24% mean inefficiency in WATRS, and a 0.41% mean inefficiency in CEP. The results indicate that, in total, flights through WATRS and CEP have approximately the same inefficiency due to off-optimal altitudes, but that blocking effects are more prevalent in WATRS. In addition, flights through WATRS are farther from their wind-optimal horizontal tracks than flights in CEP.




Marais, Purdue University.

Subject Area

Aerospace engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server