The model for the production of jets from accretion disks is relatively well established. It was demonstrated that these are open lines of a poloidal magnetic field, which extend to large distances above the disk surface into the low-density but highly magnetized disk coronae, that can extract the energy and angular momentum of the accreting matter in a form of jets. Such open lines of the magnetic field supported by the infalling gas are anchored within the relatively heavy disk, and so are co-rotating with the accreting matter. A plasma outflow can thus be centrifugally driven from the disk surface into the disk magnetosphere, along the rotating magnetic field lines, with fluid elements moving along the lines "like beads on a rotating wire". At some point, the inertia of the outflowing matter starts to play a role, winding up the magnetic field lines and forming in this way a spiral magnetic structure. The tension and pressure gradient of thus modified magnetic field assures a gradual collimation and bulk acceleration of the fluid elements in the direction perpendicular to the disk surface. As a result, a pair of bipolar jets flowing in opposite directions along the disk normal is formed, converting slowly the initially dominant magnetic flux to the bulk kinetic flux of the carried particles.
The model drafted above seems very clear, almost intuitive. Yet there are some reasons to believe it is not the mechanism primarily involved in production of relativistic jets, at least not in the case of AGN. First, quite recently we learned that the magnetic collimation of astrophysical outflows as just described is relatively inefficient in a relativistic regime. With no effective collimation provided by some additional unspecified mechanism, also the related acceleration of an outflow is deemed less efficient. Second, the whole process looks quite generic, and so one could expect that if it works for one object, it should work for all. In other words, if jets are launched from accretion disks, all the accreting black holes should be jetted. Observations tell us this is not the case, however. This may therefore imply that disk-driven outflows do not result in formation of relativistic jets, but rather only to launching of broad and at most mildly-relativistic disk winds, which indeed seem to be a general property of AGN, as already noted above. If so, the appropriate mechanism leading to the production of relativistic jets is still to be identified. Below we argue that this other mechanism of interest involves extraction of the jet energy directly from a black hole.
A black hole is characterized solely by two parameters (assuming no net electric charge), namely by its mass M and also by its angular momentum J . The mass of a black hole expressed in length units and called the "gravitational radius" is rG = GM/c2, where G is the gravitational constant, while the ratio of the black hole spin and its mass, also expressed as a length, is a = J /Mc. For a maximally rotating hole Jmax = GM2/c, i.e., a = rG. The two parameters M and J define further the event horizon (the Schwarzchild radius) as rS = rG + (rG2 - a2)1/2, and the critical distance from a singularity called the "static limit" as rC = rG + (rG2 - a2 cos2 θ)1/2, where θ is a polar coordinate. The region rS < r < rC is called the ergosphere. Every observer present in this region must rotate with the same sense as a black hole relative to a distant non-rotating frame. Still, it is allowed to exchange information between the ergosphere and the outside world. Such an exchange is not possible anymore for the region inside the event horizon, i.e. for r < rS. This means that a black hole cannot have its own magnetic field. However, a black hole can be merged into an external magnetic field supported by external currents. The crucial point is that a rotating black hole embedded in a uniform magnetic field acquires a quadrupole distribution of the electric charges with the corresponding poloidal electric field. That is a consequence of the aforementioned frame-dragging effect, forcing the magnetic field lines which are carried into the ergosphere by the accreting plasma to co-rotate with a black hole. Note that since there is no static frame in the ergosphere, there is an electric field in every frame thereby, provided the magnetic field is present. Thus, power can in principle be extracted by allowing currents to flow between the equator and poles of a spinning black hole above the event horizon, and to close at some further distance from the static limit.