It's not quite that simple; the compliance text for Part 29.865 is extensively described in AC 29.865B. Here an extract which describes (part of) the compliance check for 29.865(c)(6):
(i) In determining OEI hover performance, dynamic engine failures should be considered. Each hover verification test should begin from a stabilized hover at the maximum OEI hover weight, at the requested in-ground-effect (IGE) or OGE skid or wheel height, and with all engines operating. At this point the critical engine should be failed and the aircraft should remain in a stabilized hover condition without exceeding any rotor limits or engine limits for the operating engine(s). As with all performance testing, engine power should be limited to minimum specification power. Engine failures may be simulated by rapidly moving the throttle to idle provided a ‘needle split’ is obtained between the rotor and engine RPM.
(ii) Normal pilot reaction time should be used following the engine failure to maintain the stabilized hover flight condition. When hovering OGE or IGE at maximum OEI hover weight, an engine failure should not result in an altitude loss of more than 10 percent or four (4) feet, whichever is greater, of the altitude established at the time of engine failure. In either case, sufficient power margin should be available from the operating engine(s) to regain the altitude lost during the dynamic engine failure and to transition to forward flight.
(iii) Consideration should also be given to the time required to recover (winch up and bring aboard) the Class D external load and to transition to forward flight. This time increment may limit the use of short duration OEI power ratings. For example, for a helicopter that sustains an engine failure at a height of 40 feet, the time required to restabilize in a hover, recover the external load (given the hoist speed limitations), and then transition to forward flight (with minimal altitude loss) would likely preclude the use of 30-second engine ratings and may encroach upon the 2 ½ -minute ratings. Such encroachment into the 2 ½ - ratings is not acceptable.
(iv) For helicopters that incorporate engine driven generators, the hoist should remain operational following an engine or generator failure. A hoist should not be powered from a bus that is automatically shed following the loss of an engine or generator. Maximum two-engine generator loads should be established so that when one engine or generator fails, the remaining generator can assume the entire rotorcraft electrical load (including the maximum hoist electrical load) without exceeding approved limitations.
(v) The Rotorcraft Flight Manual (RFM) should contain information that describes the expected altitude loss, any special recovery techniques, and the time increment used for recovery of the external load when establishing maximum weights and wheel or skid heights. The OEI hover chart should be placed in the performance section of the RFM or RFM supplement. Allowable altitude extrapolation for the hover data should not exceed 2000 feet.
I have underlined the critical part; it is difficult to see how this could be extrapolated from other graphs - it requires a flight test procedure; I would guess that is should be in Part I. To my knowledge, these graphs have been produced by EC and AW for their SAR contenders.
I must say that when this was considered some time ago, the use of the wording "
an engine failure should not result in an altitude loss of more than 10 percent" did cause more than a little chuckle.
Jim