This paper examines how the planned geometry of the Einstein Telescope (ET), a next‑generation European gravitational‑wave observatory, affects how well we can measure properties of merging black holes when detectors are sometimes offline. The authors find that the nested triangular design (ET-Δ) tends to keep at least two detectors running for most of the time. That redundancy leads to tighter estimates of the source distance and the black holes’ masses during periods when the full network is not available — even in cases where the triangle design gives a lower signal‑to‑noise ratio (SNR) than the alternative two L‑shaped detectors (ET-2L). A lower SNR means the signal is less loud compared with noise.

To reach these conclusions the team built a more realistic model of detector uptime, or duty cycle, and combined it with full Bayesian parameter estimation. The duty cycle is the fraction of time a detector is producing science‑ready data. The authors model detector up and down periods as a continuous‑time Markov chain — a statistical model that tracks how many detectors are online at any moment. They fit simple exponential up and down times to real LIGO data from June–November 2023, finding average uptimes of about 5.4 hours and downtimes of about 1.6 hours. They then simulated networks of ET detectors and tested different maintenance strategies while recovering simulated gravitational‑wave signals from binary black‑hole mergers using Bayesian inference (a method that gives probability distributions for source properties and their uncertainties).

A key operational difference they tested is maintenance scheduling. For ET-Δ a rotating maintenance plan has been proposed: keep two of the three V‑shaped detectors online while the third is serviced. That uses the triangle’s redundancy and keeps multi‑detector coverage high. For ET-2L, which consists of two separated L‑shaped instruments, maintenance is likely to be synchronized in the usual way used today. Under those assumptions the Markov‑chain model predicts ET-Δ spends most observing time with two or three detectors online, while ET-2L more frequently has only a single detector operating. With two or more detectors the network can better locate sources on the sky and separate parameters that are otherwise mixed together, so distance and source‑frame component masses (the masses of the black holes as measured in their own rest frame) are recovered more tightly.