X-ray cavities are the visible receipts of cluster-core energy injection: bubble-shaped deficits in the intracluster medium's X-ray emission, carved by relativistic plasma and conventionally attributed to jets from the central active galactic nucleus. The receipts and the income do not reconcile. The mechanical energy required to excavate observed cavity systems frequently exceeds the central AGN's luminosity and plausible duty cycle by factors of several; cavities appear too large, too numerous, and too old for the current outburst to have made them; and some clusters hold multiple generations of cavities with no corresponding record of AGN activity at the required epochs (McNamara and Nulsen 2007; Fabian 2012).
The mismatch matters doubly because the standard model needs the AGN twice over: the same feedback invoked to carve cavities is the designated regulator preventing runaway cooling in cluster cores, so an energy ledger that does not balance undermines both the cavities' origin and the cooling-flow solution built on it. The repair options, AGN more powerful and longer-lived than their observed luminosities suggest, or an unidentified additional energy source, either strain the engine physics or concede the inventory is incomplete.
The standing is an accounting dispute at the heart of cluster feedback theory, with every deep Chandra cavity census adding receipts the central engine cannot cover.