• Energy Research

As a result of the increasing integration of Renewable Energy Source (RES), maintenance of the balance between supply and demand in the power system is more challenging because of RES’s intermittency and uncontrollability. The smart control of demand is able to contribute to the balance by providing the grid frequency response. This paper uses the industrial Melting Pot (MP) loads as an example. A thermodynamic model depicting the physical characteristics of MPs was firstly developed based on field measurements carried out by Open Energi. A distributed control was applied to each MP which dynamically changes the aggregated power consumption of MPs in proportion to changes in grid frequency while maintaining the primary heating function of each MP. An aggregation of individual MP models equipped with the control was integrated with the Great Britain (GB) power system models. Case studies verified that the aggregated MPs are able to provide frequency response to the power system. The response from MPs is similar but faster than the conventional generators and therefore contributes to the reduction of carbon emissions by replacing the spinning reserve capacity of fossil-fuel generators. Through the reviews of the present balancing services in the GB power system, with the proposed frequency control strategy, the Firm Frequency Response service is most beneficial at present for demand aggregators to tender for. All studies have been conducted in partnership between Cardiff University, Open Energi London – Demand Aggregator, and National Grid – System Operator in GB to ensure the quality and compliance of results.

The highly penetrated distributed generators (DGs) aggravate the voltage violations in active distribution networks (ADNs). The coordination of various regulation devices such as on-load tap changers (OLTCs) and DG inverters can effectively address the voltage issues. Considering the problems of inaccurate network parameters and rapid DG fluctuation in practical operation, multi-source data can be utilized to establish the data-driven control model. In this paper, a data-driven coordinated voltage control method with the coordination of OLTC and DG inverters on multiple time-scales is proposed without relying on the accurate physical model. First, based on the multi-source data, a data-driven voltage control model is established. Multiple regulation devices such as OLTC and DG are coordinated on multiple time-scales to maintain voltages within the desired range. Then, a critical measurement selection method is proposed to guarantee the voltage control performance under the partial measurements in practical ADNs. Finally, the proposed method is validated on the modified IEEE 33-node and IEEE 123-node test cases. Case studies illustrate the effectiveness of the proposed method, as well as the adaptability to DG uncertainties.

La estimación estatal juega un papel vital para monitorear, observar y comprender la red combinada de calor y electricidad. En este documento, se presenta un marco híbrido para estimar con precisión los estados del sistema de la red de distribución eléctrica y la red de calor, utilizando las mediciones limitadas no redundantes obtenidas del control de supervisión y la adquisición de datos y los sistemas avanzados de infraestructura de medición. El marco híbrido presentado implica dos pasos, a saber, la previsión del estado y la estimación del estado. El pronóstico de estado utiliza una red neuronal profunda para pronosticar los estados del sistema cada quince minutos, mientras que estos estados pronosticados son utilizados por el estimador híbrido, que utiliza un filtro Kalman extendido robusto para estimar los estados del sistema con la ayuda de ambos conjuntos de datos correspondientes al control de supervisión y la adquisición de datos y los sistemas de infraestructura de medición avanzada, a intervalos de una hora. El marco propuesto no se basa completamente en el modelo del sistema en diferentes instantes. La efectividad del método se valida a través de comparaciones exhaustivas con estudios de simulación realizados utilizando el sistema de prueba Barry Island, Reino Unido. Se observa un rendimiento satisfactorio incluso con la presencia de datos incorrectos en las mediciones. L'estimation de l'état joue un rôle essentiel pour surveiller, observer et comprendre le réseau combiné de chaleur et d'électricité. Dans cet article, un cadre hybride est présenté pour estimer avec précision les états du système du réseau de distribution électrique et du réseau de chaleur, en utilisant les mesures non redondantes limitées obtenues à partir des systèmes de contrôle de surveillance et d'acquisition de données et d'infrastructure de mesure avancée. Le cadre hybride présenté comprend deux étapes, à savoir la prévision de l'état et l'estimation de l'état. La prévision d'état utilise un réseau neuronal profond pour prévoir les états du système toutes les quinze minutes, tandis que ces états prévus sont en outre utilisés par l'estimateur hybride, qui utilise un filtre de Kalman étendu robuste pour estimer les états du système à l'aide des deux ensembles de données correspondant au contrôle de supervision et à l'acquisition de données et aux systèmes d'infrastructure de mesure avancés, à intervalle horaire. Le cadre proposé ne repose pas entièrement sur le modèle du système à différents instants. L'efficacité de la méthode est validée par des comparaisons approfondies avec des études de simulation réalisées à l'aide du système de test Barry Island, Royaume-Uni. Des performances satisfaisantes sont observées même avec la présence de mauvaises données dans les mesures. State-estimation plays a vital role to monitor, observe and understand the combined heat and electric network. In this paper, a hybrid framework is presented to accurately estimate the system states of electric distribution network and heat network, using the limited non-redundant measurements obtained from supervisory control and data acquisition and advanced metering infrastructure systems. The presented hybrid framework involves two steps, namely, the state-forecasting and the state-estimation. The state-forecasting uses a deep neural network to forecast the system states at every fifteen minutes interval, while these forecasted states are further used by the hybrid estimator, which uses a robust extended Kalman filter to estimate the system states with help of both datasets corresponding to supervisory control and data acquisition and advanced metering infrastructure systems, at hourly interval. The proposed framework does not completely rely on the system model at different instants. The effectiveness of the method is validated through thorough comparisons with simulation studies carried out using the Barry Island test system, United Kingdom. Satisfactory performance is observed even with the presence of bad data in the measurements. يلعب تقدير الحالة دورًا حيويًا في مراقبة وملاحظة وفهم الشبكة الحرارية والكهربائية المدمجة. في هذه الورقة، يتم تقديم إطار عمل هجين لتقدير حالات نظام شبكة التوزيع الكهربائية والشبكة الحرارية بدقة، باستخدام القياسات المحدودة غير الزائدة عن الحاجة التي تم الحصول عليها من التحكم الإشرافي والحصول على البيانات وأنظمة البنية التحتية المتقدمة للقياس. يتضمن الإطار الهجين المقدم خطوتين، وهما التنبؤ بالحالة وتقدير الحالة. يستخدم التنبؤ بالحالة شبكة عصبية عميقة للتنبؤ بحالات النظام كل خمسة عشر دقيقة، في حين يتم استخدام هذه الحالات المتوقعة بشكل أكبر من قبل المقدر الهجين، والذي يستخدم مرشح كالمان الممتد القوي لتقدير حالات النظام بمساعدة كل من مجموعات البيانات المقابلة للتحكم الإشرافي والحصول على البيانات وأنظمة البنية التحتية المتقدمة للقياس، على فترات كل ساعة. لا يعتمد الإطار المقترح بشكل كامل على نموذج النظام في لحظات مختلفة. يتم التحقق من فعالية الطريقة من خلال مقارنات شاملة مع دراسات المحاكاة التي أجريت باستخدام نظام اختبار جزيرة باري، المملكة المتحدة. يلاحظ الأداء المرضي حتى مع وجود بيانات سيئة في القياسات.

The development of the North Sea Supergrid is closely linked to three main topological ideas: (i) Business as Usual; (ii) Local Coordination; and (iii) Fully Integrated. In this paper, the three proposed topologies were simplified, analysed and compared for power export from offshore wind farms unto onshore grids. A review on the state-of-the-art HVAC and HVDC offshore transmission technologies was carried out. The national offshore network development strategies of the six countries surrounding the North Sea were summarised. The number of HVAC and HVDC assets required for all three topologies were calculated. It was evident from calculations that, by 2030, the Fully Integrated Topology will achieve approximately 8% asset savings than the Local Coordination Topology.

AbstractThe carbon constrained design of energy supply infrastructure for new build schemes was investigated. This was considered as an optimization problem with the objective of finding the mix of on-site energy supply technologies that meet green house gas emissions targets at a minimum build cost to the developer. An integrated design tool was developed by combining a social cognitive optimisation solver, an infrastructure model and a set of analysis modules to provide the technical design, the evaluation of greenhouse gas emissions and the financial appraisal for the scheme. The integrated design tool was applied to a new build scheme in the UK with a 60% target reduction of regulated emissions. It was shown that the optimal design and corresponding cost was sensitive to the year of build completion and to the assumptions applied when determining the emissions intensity of the marginal central generators.

As the integration of wind generation displaces conventional plants, system inertia provided by rotating mass declines, causing concerns over system frequency stability. This paper implements an advanced stochastic scheduling model with inertia-dependent fast frequency response requirements to investigate the challenges on the primary frequency control in the future Great Britain electricity system. The results suggest that the required volume and the associated cost of primary frequency response increase significantly along with the increased capacity of wind plants. Alternative measures (e.g. electric vehicles) have been proposed to alleviate these concerns. Therefore, this paper also analyses the benefits of primary frequency response support from electric vehicles in reducing system operation cost, wind curtailment and carbon emissions.