Usually, the Traveller M-Drive violates conservation of energy. But what if it did not? And how could that be, given the energy consumptions (and the zero-reaction-mass assumption) from Traveller starships' maneuver drives?
If we simply assume that the Traveller maneuver drive has a way to anchor the ship to spacetime and thus magically allows the ship to use the whole universe as its reaction mass (much like we all use the planet Earth as our reaction mass when jumping), then you'll need only negligible power for accelerating the reaction mass, even though this reaction mass is 5*10^52 kg. How?
When starting, we "accelerate the universe" to a speed of less than 1 Planck length per second. For example, for accelerating a 100-ston-ship to 11 m/s, you'd have to accelerate the universe to 2.2*10^(-47) meters per second. (A Planck length would be 1.6*10^(-35) meters). This requires a total of
0,5*5*10^52 kg * [2.2*10^(-47)]^2=
1,21*10^(-41) Joules.
In other words, not an amount of energy worth mentioning. Really. :-)
Now, in order to reach a given speed, we will need energy according to the following formula:
[2*(Energy capacity of M-Drive in Watts)*(time)/(ship's mass in kg)]^(0.5)
[(speed)^2*(ships's mass in kg)/(2*energy capacity of M-Drive in watts)]
Speed (m/s) Time required (s) Time required (h) Time required (d) 0 0 0 0 10 0,33 0 0 20 1,31 0 0 50 8,18 0 0 100 32,73 0,01 0 200 130,91 0,04 0 500 818,18 0,23 0,01 1,000 3,272,73 0,91 0,04 2,000 13,090,91 3,64 0,15 5,000 81,818,18 22,73 0,95 10,000 327,272,73 90,91 3,79 20,000 1,309,090,91 363,64 15,15 50,000 8,181,818,18 2,272,73 94,7 100,000 32,727,272,73 9,090,91 378,79
However, since actual energy consumption would depend on the current speed in relation to when the ship's drive was activated first, the bookeeping required for this would be forbidding. So let's forget the whole idea. :-)