Investigation of building energy flexibility at cluster level for a promising energy flexibility market / submitted by Tuğçin Kırant-Mitić

Kırant-Mitić, Tuğçin

Buildings are becoming prosumers with the electrification of heat and mobility and the spread of rooftop photovoltaic and battery storage. Their operation should follow price and CO₂ signals while prioritizing distribution-level feasibility, including transformer capacity, voltage limits, and node-specific import/export caps. This cumulative dissertation investigates how building-side demand flexibility can be modeled and controlled so that buildings deliver bidirectional value, more specifically, lower cost and emissions for end-users and reliable flexibility for the power grid without compromising comfort. This dissertation focuses on flexibility at the distribution system level. Specifically, it examines how building´s heating, cooling, air-conditioning and ventilation systems (i.e., heat pumps, thermal storages) and distributed energy resources (i.e., electrical energy storage and photovoltaic) can provide location and time-specific flexibility to manage both market-oriented operation and grid-oriented operation within low voltage network. The work (i) characterizes distribution system operator needs in the frame of flexibility, (ii) develops and simulates control strategies for buildings such as ranging from price-driven responses to explicit congestion signals, (iii) evaluates performance on realistic German building clusters, and (iv) quantifies benefits in terms of reduced constraint violations and operational cost/indoor thermal comfort impacts for end users. Interactions with transmission system operators are treated as boundary conditions; the primary objective is operational feasibility and value for distribution system operator, ensuring flexibility is applicable, verifiable, and aligned with local network constraints. In this context, the dissertation has four main contributions. First (building-gridi nteraction signal evaluation), it develops a joint penalty signal that combines economic (day-ahead price) and environmental (average CO₂-equivalent intensity) drivers into a single control objective. The analysis compares the composite objective with single-signal control and presents demand-flexibility results for cases where price and CO₂ signals align or conflict. Second, in the single-residential-building context, a simulation case study with heuristic control demonstrates that a small dwelling can deliver grid-oriented and market-oriented operation simultaneously, achieving grid support and cost reduction. Third (single non-residential building operation), it introduces a rule-based predictive control architecture implemented on a building with thermally activated building systems. In a co-simulation workflow that couples a white-box building energy performance model with a numerical computing engine, a horizon controller uses day-ahead prices and distribution system operator constraints to schedule preheating/precooling and charge thermal storage in the building mass,while applying comfort via slab and zone temperature limits. Fourth (cluster operation), two transformer-aware case studies in Germany and Switzerland quantify how building-grid interaction signals shape flexibility at the multi-building scale, showing that local constraints, both load-driven and generation-driven, govern performance. Within a calibrated co-simulation framework, the study analyzes cluster operation under price and CO₂eq. intensity signals while enforcing transformer limits, thereby delivering transformer-aware demand-side flexibility. Overall, the research analyzes both grid-oriented and market-oriented operation, including a sequential “grid-first, market-second” strategy: grid constraints are treated as primary, and economic/environmental objectives are optimized subject to those constraints. The evaluation presents both single-building and multi-building scales: (i) residential and non-residential single buildings and (ii) clusters in different countries (Germany and Switzerland), where transformer-aware scheduling is shown. Methodologically, the work introduces a new control algorithm „Rule-basedpredictive control“ as an alternative to simple rule-based control and computationally heavy control models. Rule-based predictive control runs in a co-simulation workflow with a calibrated white-box building model, a horizon scheduler, and distribution-level limits. The results show that the grid-first, market-second approach reliably prevents local violations while still achieving cost and CO₂ benefits. Moreover, rule-based predictive control captures most of the expected performance with only a fraction of the computational effort, and at the cluster scale, transformer-aware control reduces coincident peaks and curtailment, outperforming single-signal baselines. Parts of this work were undertaken within the InFleX project (EFRE 2014–2020) and the International Energy Agency’s EBC Annex 82 - Energy Flexible Buildings Towards Resilient Low Carbon Energy Systems programme, which provided knowledge exchange, shared scenarios, and cross-case comparability. In turn, the dissertation’s tools and findings contribute to guidance on assessing and delivering verifiable flexibility.

Detailsuche

Bibliotheken

Projekt

Impressum

Datenschutz

Titelaufnahme