Meh, its much more boring than all that. Orbital speeds at 1000km up aren't that fast (I'd say more like 5km/s impact speed)
Remember that you need to consider the kinetic energy from the thousand (well, about 900 km or so till atmosphere starts to decelerate it measurably) kilometre descent as well.
The only way an object from high earth orbit would suddenly fall to earth is if it's velocity suddenly decreased enough to bring the perigee inside the planet's surface sphere. For sake of simplicity, let's say that the object is at roughly 3600 kilometres altitude spherical orbit (10000 km orbital radius). At this situation, orbital velocity is slightly less than 6000 m/s (5900 m/s to be exact).
Suddenly, the ship's thrusters grossly malfunction due to battle damage or Carl's lunch dripping blood over control panels and causing a short, whatever. The thrusters fire up in retrograde direction, which causes the orbital velocity to decrease. The required velocity change to bring the perigee within the planet's surface(orbit intersects planet ie. collision occurs) is about 700 m/s from this particular orbit; when orbital velocity at 10000 km apogee is less than approximately 5200 m/s, the orbit's perigee sinks below the planet's surface. This means a very shallow impact angle, however; for greater impact angle the delta v needs to be larger, and correspondingly the impact velocity will be slightly less.
However, from this point the ship's velocity increases as it's falling towards the planet. Crossing the semi-minor axis point, the orbital velocity is already at about 7000 m/s and increasing.
At the beginning of atmospheric entry, velocity is about 8200 m/s which is scarily close to my initial estimation of 8 km/s which I presented
here 
. I flew a delta glider in Orbiter to test the numbers (it was faster than calculating and more fun) and a Sathanas would have the same trajectory up until the point of atmospheric entry, after which it wouldn't be significantly slowed down by atmosphere like my puny little glider - it would hit the ground at around 8 km/h with this kind of start-up orbital situation.
...and the sathanas is going to be more like a 3km sphere. Half of it is long arms that you could fold back to stuff it in a box. Also, it is going to have internal space and lots of it. I figure it (like most all ships space and otherwise) would float if it landed in the ocean, so I'd use a density around 1000kg/m^3. Also if its coming down from orbit, it is going to hit at a very glancing angle most likely, so that gives the following impact:
Sounds pretty much the same as my rationales for my values earlier.
Of course, nothing forces the Sathanas to settle on an orbit to begin with - it could be on hyperbolic trajectory around Sun and hit the Earth at any given speed within reasonable limits. Something like 100 km/s wouldn't sound too impossible for such situation if you want to maximize the destruction...
