Let\'s Experiment: roughly duplicate http://en.wikipedia.org/wiki/Bi-elliptic_transfer#Example. Experiment 1, in which Bi-Elliptic should barely be more efficient. Radius of Kerth: 600km. Starting altitude: 100km. Initial radius, then, is 700km. Desired final circular orbit: 10000km altitude. This is 10600 radius, or ~15x our starting radius. Turnover altitude: 28000km. This is a radius of 28600, or ~41x our starting radius. We\'ll Contrast this with a Hohmann transfer directly to 10000km. [table] [tr][td]Burn [/td][td]Hohmann dV [/td][td]Bi-elliptic dV[/td][/tr] [tr][td]1[/td][td]830[/td][td]632 or 892[/td][/tr] [tr][td]2[/td][td]374[/td][td]157[/td][/tr] [tr][td]3[/td][td]-[/td][td]114[/td][/tr] [tr][td]Total[/td][td]1204[/td][td]903 or 1163[/td][/tr] [/table] The struck out values above are given by the program. I believe the first should be 892 (the same as the first burn of a Hohmann transfer from 100km to 28000km). The total, then, also changes, obviously. Note that, as expected, the Bi-elliptic is slightly more efficient in fuel than the Hohmann. This mirrors the Wikipedia article. Experiment 2: In which Bi-elliptic should be less efficient Now, let\'s experiment again, with a destination orbit that\'s lower than the magical 11.8 ratio. Let\'s say 5000 km (for a radius of 5600=8x our initial radius), using the same turnover altitude. [table] [tr][td]Burn [/td][td]Hohmann dV [/td][td]Bi-elliptic dV[/td][/tr] [tr][td]1[/td][td]749[/td][td]632 or 892[/td][/tr] [tr][td]2[/td][td]420[/td][td]104[/td][/tr] [tr][td]3[/td][td]-[/td][td]215[/td][/tr] [tr][td]Total[/td][td]1168[/td][td]951 or 1211[/td][/tr] [/table] If the value for the transfer orbit injection in the bi-elliptic calculator is correct, this still takes less delta-v. This can\'t be true for a ratio of the major axes that\'s only 8x. However, if we use the value for the same burn obtained using the Hohmann calculator (the same 892 as before), we see that the bi-elliptic is less efficient. I think it\'s clear that the program is just giving the wrong values for delta v and final v for the 'transfer orbit injection' on the bi-elliptic page (and consequently also reporting the wrong total delta v needed). Really, the giveaway is the fact that the sum of the initial and the deltav doesn\'t produce the final. It\'s just clearly in error.