The central engine of gamma-ray bursts ( GRBs ) is poorly constrained .
There exist two main candidates : a fast-rotating black hole and a rapidly spinning magnetar .
Furthermore , X-ray plateaus are widely accepted to be the energy injection into the external shock .
In this paper , we systematically analyze the Swift /XRT light curves of 101 GRBs having plateau phases and known redshifts ( before 2017 May ) .
Since a maximum energy budget ( \sim 2 \times 10 ^ { 52 } erg ) exists for magnetars but not for black holes , this provides a good clue to identifying the type of GRB central engine .
We calculate the isotropic kinetic energy E _ { K,iso } and the isotropic X-ray energy release E _ { X,iso } for individual GRBs .
We identify three categories based on how likely a black hole harbors a central engine : ’ Gold ’ ( 9 out of 101 ; both E _ { X,iso } and E _ { K,iso } exceed the energy budget ) , ’ Silver ’ ( 69 out of 101 ; E _ { X,iso } less than the limit but E _ { K,iso } is greater than the limit ) , and ’ Bronze ’ ( 23 out of 101 , the energies are not above the limit ) .
We then derive and test the black hole parameters with the Blandford-Znajek mechanism , and find that the observations of the black hole candidate ( ’ Gold ’ + ’ Silver ’ ) samples are consistent with the expectations of the black hole model .
Furthermore , we also test the magnetar candidate ( ’ Bronze ’ ) sample with the magnetar model , and find that the magnetar surface magnetic field ( B _ { p } ) and initial spin period ( P _ { 0 } ) fall into reasonable ranges .
Our analysis indicates that if the magnetar wind is isotropic , a magnetar central engine is possible for 20 % of the analyzed GRBs .
For most GRBs , a black hole is most likely operating .