You cannot have a simple spread footing with e > L/2. This puts the resultant load off the footing, which is unacceptable. Either make the footing bigger, or use a strap beam to resist the moment.
It would depend on the situation. I believe that most codes would require your loading condition to produce a conservative result.
For example, when analyzing the external stability of a MSE retaining wall, live loads supported directly on the MSE fill are ignored when calculating the overturning and sliding resistance. Live loads on the retained soil are included in these cases, as they will tend to increase the eccentricity or driving lateral force.
Self-weight of the foundation should be included in the eccentricity calculation, as should the weight of the soil directly above the footing, but not the column/wall stem.
I would explicitly ignore any "passive pressure" other than self-weight provided by soil directly above a foundation. Watch out for assuming soil above the foundation if structural loads will be applied before backfilling.
I would have to disagree with Mr. Atkins. While it may be true in building construction, we use spread footings for high overturning moments all the time in substations. The RUS Bulletin 1724E-300 shows some examples. See:
MikeDB, you can use the weight of the soil and the selfweight of the footing to increase P, but e = M/P MUST be less than one-half of the width of the footing, or the vertical forces cannot be balanced (because the resisting soil pressure is only under the footing).
I looked in the RUS Bulletin you mentioned (good reference by the way, I wasn't aware of this one, not being in the electrical industry). If the eccentricity is equal to half the footing width, e=b/2, then the safety factor is equal to 1.0 for overturning (substitute Eq. 8-18 and 8-20 into 8-19). However the bulletin states that a safety factor between 1.5 and 2.0 is adequate. This agrees with what DaveAtkins stated above and with which I concur; you cannot safely use a spread footing where the eccentricity is greater than half the footing width. If you cannot make the footing wider, then you should consider piles or drilled shafts.
I stand corrected. I was thinking of the eccentricity of the applied load (moment at top of footing divided by shear at top of footing) being well outside the footing.
