Demand destruction is the reduction in consumption caused by high prices rather than by a change in activity. It is the ultimate balancing mechanism: if supply cannot rise and inventory cannot fall further, price must rise until enough demand goes away.
Short-run price elasticity of oil demand is low. People still drive to work, aircraft still fly their schedules, and industrial processes still need feedstock, so a large price rise produces only a small immediate volume response. That low elasticity is exactly why prices have to move so far to clear a supply shock: a small quantity adjustment requires a large price adjustment.
Long-run elasticity is considerably higher, because the adjustment happens through the capital stock rather than through behaviour. Vehicles get replaced with more efficient ones, industrial processes get redesigned, and substitution to other fuels becomes worth its switching cost. That takes years, and it does not reverse when the price falls, which is why a price spike leaves a permanent mark on the demand trajectory.
Substitution deserves a specific mention because it is the fastest channel. Power generators that can switch between gas and oil products do so on relative price, and that switching is genuinely rapid where the dual-fuel capability exists. It is a small share of total demand but a disproportionate share of the short-run elasticity.
The asymmetry with seasonality is worth holding onto. Seasonal demand changes are large, predictable and reverse. Destruction is smaller, slower, harder to see, and partly permanent. Confusing a seasonal decline for destruction, or the reverse, is a common and expensive error.