Reversing the comparison makes the planning problem legible. If 1.7% a year is more than ten times the pace of the preceding fifteen years [1], the earlier trend ran at roughly 0.17% a year or below [12], which is flat for the purposes of anyone sizing a maintenance budget or a replacement cycle. Compounded, five years at 1.7% is about 8.8% more load riding the same conductors [13]. That is not a large number in isolation, and it is a large number if the staffing plans and spares inventories underneath it were written for a system nobody expected to grow.
At the system level, 8.8% over five years is small next to the swings utilities absorb from weather every year. The load, though, does not arrive at the system level. The projection that data center demand nearly triples by 2030 implies, if its baseline is roughly the present, about 25% compound annual growth in a single class of customer [14], and that customer connects at a named substation rather than across a state.
Restoration speed separates from available megawatts because the two answers come from different parts of a utility. Capacity is a planning and interconnection question, settled years ahead in studies and queue positions. Restoration is a question of crews, spares and physical access, settled on the day. The contributor's argument is that reactive maintenance was tolerable while demand stayed flat and no longer is [8], and that the mismatch already shows up as conductor sag and overheated transformers under today's loads [16].
The evidence deserves two caveats. The piece is a contributor column, and Entrepreneur states the opinions are the contributor's own [11]; its load-bearing demand figure is attributed to federal energy forecasts without naming the forecast or the agency [1]. Its remedy section argues for specialized high-voltage line repair gear, including insulated aerial platforms and modular rigs said to cut repair windows from days to hours by avoiding full feeder de-energization [5]. The direction of the demand claim is credible, and the days-to-hours figure is the shape of claim a supplier makes, which this record does not let us test.
Sequencing is what should govern behaviour this quarter. New transmission into terrain utilities have rarely serviced, which is what remote wind and solar sites require [9], is a decade-scale build, and no site chosen in the next ninety days will benefit from it. What a tenant can still buy is written into the lease and the interconnection agreement, or installed on the property, and both are priced against the network as it stands. The trade is between paying for that hedge during site selection and learning its price on the day the substation serving you fails.
