Most multifamily housing in Sweden is heated by district heating, and renovating. The stock built during the Million Program is central to national and EU efficiency targets. The worth of such a renovation is usually judged by the saving its new components are calculated to deliver. This thesis examines a case where that expectation did not hold. At Ekebacken 1, a district heated estate of identical blocks built in 1981 in Ljungby, the windows, doors and thermostatic valves were replaced in phases, work that should have cut the annual district heating demand by a clear margin. The property meter barely moved, and most of the expected saving never appeared.The study explains this gap by testing four candidate causes against measured evidence, drawing only on operational data the property owner already keeps, brought together through a per block energy signature method.The four candidates are, fabric and heating network loss, windows and doors installation, rebound effect and residents behavior, and hydraulic imbalance in heating system.The building fabric is found to be uniform across the estate, and the block-to-block variation tracks the flow that each block draws from the shared circuit rather than its distance from the substation. The substation temperature difference is below its design values leads to estate draws more water than it needs and returns it too warm. Taken together, these signals are consistent with hydraulic imbalance in the un renovated secondary circuit. The envelope improvement is genuine, but an unbalanced circuit distributes the heat unevenly, so the reduced demand does not register at the meter.This thesis argues that the saving expected from the renovation failed to appear at the meter not because the envelope work was deficient, but because the imbalanced secondary circuit distributes the heat by hydraulic resistance, in the meantime, the lost saving can be recovered by balancing the circuit.