Statistical load and generation modelling for long term studies of low voltage networks in presence of sparse smart metering data

Toubeau, Jean‐François; Hupez, Martin; Klonari, Vasiliki; Grève, Zacharie De; Vallée, François

doi:10.1109/iecon.2016.7793228

Cited by 6 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The promising outcomes of the work pave the way towards more advanced strategies, such as the extension to a decentralized approach using a multi-agent formulation (that would prevent the single point of failure of the centralized framework). Similarly, extending the framework to partially observable networks (where the state of the system is not fully known [40]) also offers a valuable area of research for system operators.…”

Section: Discussionmentioning

confidence: 99%

Deep Reinforcement Learning-Based Voltage Control to Deal with Model Uncertainties in Distribution Networks

et al. 2020

Self Cite

View full text Add to dashboard Cite

This paper addresses the voltage control problem in medium-voltage distribution networks. The objective is to cost-efficiently maintain the voltage profile within a safe range, in presence of uncertainties in both the future working conditions, as well as the physical parameters of the system. Indeed, the voltage profile depends not only on the fluctuating renewable-based power generation and load demand, but also on the physical parameters of the system components. In reality, the characteristics of loads, lines and transformers are subject to complex and dynamic dependencies, which are difficult to model. In such a context, the quality of the control strategy depends on the accuracy of the power flow representation, which requires to capture the non-linear behavior of the power network. Relying on the detailed analytical models (which are still subject to uncertainties) introduces a high computational power that does not comply with the real-time constraint of the voltage control task. To address this issue, while avoiding arbitrary modeling approximations, we leverage a deep reinforcement learning model to ensure an autonomous grid operational control. Outcomes show that the proposed model-free approach offers a promising alternative to find a compromise between calculation time, conservativeness and economic performance.

show abstract

Section: Discussionmentioning

confidence: 99%

Deep Reinforcement Learning-Based Voltage Control to Deal with Model Uncertainties in Distribution Networks

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…The four additional years are simply obtained by feeding the existing years into a Seasonal AutoRegressive Moving Average (SARMA) model that is used to generate new representative time trajectories. 36 The general statistical information on price scenarios are shown in Table 2. Likewise, the hourly needs in terms of upward secondary reserve for a typical week are depicted in Figure 7, where we observe the high variability of the balancing needs.…”

Section: Case Studymentioning

confidence: 99%

“…The non-linear charge and discharge curves (26), are approximated using two piecewise linear segments. The resulting MILP problem (11)- (36), with N ω = 10 scenarios composed of N t = 8760 hours, is characterized by 4 029 615 constraints, 1 138 802 continuous variables, and 525 600 binary variables. It is implemented in Julia/JuMP, and solved with Gurobi 8.1.1, on a 16 GB-RAM computer clocking at 3.40 GHz.…”

Section: Case Studymentioning

confidence: 99%

“…To that end, we rely on 6 years of data (from 2012 to 2017) that are directly fed to the stochastic program, along with four artificially‐generated years, in order to have a total of 10 typical years (which thus correspond to 3650 representative days). The four additional years are simply obtained by feeding the existing years into a Seasonal AutoRegressive Moving Average (SARMA) model that is used to generate new representative time trajectories 36 . The general statistical information on price scenarios are shown in Table 2.…”

Section: Case Studymentioning

confidence: 99%

See 1 more Smart Citation

Sizing of underwater gravity storage with solid weights participating in electricity markets

Toubeau

Ponsart

Stevens³

et al. 2020

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

This paper investigates the techno-economic feasibility of the innovative concept of gravity energy storage, where heavy weights are raised and lowered in a water environment. Such eco-friendly systems can be implemented in existing flooded pits or quarries, by leveraging the important depth of these cavities. Moreover, in addition to their long lifetime, they have no visual impact on the landscape, and offer a lot of flexibility to the power system. In this work, we firstly present an analytical study of the storage solution, which allows deriving tractable mathematical constraints describing its operation, such as its nonlinear speed-power curves in both charge and discharge modes. These constraints are then integrated into the investment strategy of a merchant unit that seeks to maximize its profit by jointly participating in the energy and secondary reserve markets. The model is formulated within a stochastic framework to ensure robustness of sizing decisions in view of future market uncertainties. Results from a practical case study (on a natural cavity of 200 m) show that underwater gravity storage is a cost-efficient technology that offers payback periods of less than 10 years, mainly due to its intrinsic low capital costs estimated at around 100 €/kWh. List of Symbols and Abbreviations: n blocks , Number of installed blocks; n gm , Number of installed geared induction machines; φ CAPEX , Total investment costs of the storage system, €; φ OPEX ω,t , Total profit from storage operation in scenario ω at time t, €; n blocks ω,t , Number of blocks stored within the storage system in scenario ω at time t; p dis,res ω,t , p ch,res ω,t , Actual output power in discharge (dis) and charge (ch) modes after the real-time activation of reserves in scenario ω at time t, MW; p dis ω,t , p ch ω,t , Output power in discharge (dis) and charge (ch) modes committed in the energy market in scenario ω at time t, MW; p single,Á,res ω,t , Output power per induction machine in discharge (dis) and charge (ch) modes after the real-time activation of reserves in scenario ω at time t, MW; res dis ω,t ,res ch ω,t , Reserve capacity allocated in the upward secondary reserve market in discharge (dis) and charge (ch) mode in scenario ω at time t, MW; v dis ω,t ,v ch ω,t , Aggregated block speed (at the UGES level) in discharge (dis) and charge (ch) mode in scenario ω at time t, m/s; v single,Á ω,t , Block speed per induction machine in discharge (dis) and charge (ch) mode in scenario ω at time t, m/s; z dis ω,t ,z ch ω,t , Binary variables indicating the discharge (dis) and charge (ch) status in scenario ω at time t; Δ t , Optimization step, 1 hour; η dis , η ch , Efficiency of the storage system in discharge (dis) and charge (ch) modes; κ ω,t ∈ [0,1], Activation rate of the upward reserve; λ act ω,t , Price for activation of balancing reserves in scenario ω at time t, €/MWh; λ EN ω,t , Electricity price in the energy-only market in scenario ω at time t, €/MWh; λ res ω,t , Price for availability of reserve capacity in scenario ω at time ...

show abstract

“…More advanced techniques try to capture the full diversity by clustering the consumers based on the available attributes and sampling profiles from the cluster after fitting a probability distribution [4] or training a Markov model [5] on each cluster. Bottom-up approaches model the time series consumption data of a consumer based on occupant behaviour and electrical appliance usage data [6].…”

Section: Introductionmentioning

confidence: 99%

Scenario generation of residential electricity consumption through sampling of historical data

Soenen

Yurtman

Becker

et al. 2023

Sustainable Energy, Grids and Networks

View full text Add to dashboard Cite

Statistical load and generation modelling for long term studies of low voltage networks in presence of sparse smart metering data

Cited by 6 publications

References 20 publications

Deep Reinforcement Learning-Based Voltage Control to Deal with Model Uncertainties in Distribution Networks

Deep Reinforcement Learning-Based Voltage Control to Deal with Model Uncertainties in Distribution Networks

Sizing of underwater gravity storage with solid weights participating in electricity markets

Scenario generation of residential electricity consumption through sampling of historical data

Contact Info

Product

Resources

About