No abstract

The UK National Grid has placed increased emphasis on the development of Demand Side Response (DSR) tariff mechanisms to manage load at peak times. Refrigeration systems, along with HVAC, are estimated to consume 14% of the UK’s electricity and could have a significant role for DSR application. However, characterized by relatively low individual electrical loads and massive asset numbers, multiple low power refrigerators need aggregation for inclusion in these tariffs. In this paper, the impact of the Demand Side Response (DSR) control mechanisms on food retailing refrigeration systems is investigated. The experiments are conducted in a test-rig built to resemble a typical small supermarket store. The paper demonstrates how the temperature and pressure profiles of the system, the active power and the drawn current of the compressors are affected following a rapid shut down and subsequent return to normal operation as a response to a DSR event. Moreover, risks and challenges associated with primary and secondary Firm Frequency Response (FFR) mechanisms, where the load is rapidly shed at high speed in response to changes in grid frequency, is considered. For instance, measurements are included that show a significant increase in peak inrush currents of approx. 30% when the system returns to normal operation at the end of a DSR event. Consideration of how high inrush currents after a DSR event can produce voltage fluctuations of the supply and we assess risks to the local power supply system.

Traditional techniques for balancing long, flexible, high-speed rotating shafts are inadequate over a full range of shaft\ud speeds. This problem is compounded by limitations within the manufacturing process, which have resulted in increasing\ud problems with lateral vibrations and hence increased the failure rates of bearings in practical applications. There is a need\ud to develop a novel strategy for balancing these coupling shafts that is low cost, robust under typically long-term operating\ud conditions and amenable to on-site remediation. This paper proposes a new method of balancing long, flexible couplings\ud by means of a pair of balancing sleeve arms that are integrally attached to each end of the coupling shaft. Balance\ud corrections are applied to the free ends of the arms in order to apply a corrective centrifugal force to the coupling shaft\ud in order to limit shaft-end reaction forces and to impart a corrective bending moment to the drive shaft that limits shaft\ud deflection. The aim of this paper is to demonstrate the potential of this method, via the mathematical analysis of a plain,\ud simply supported tube with uniform eccentricity and to show that any drive shaft, even with irregular geometry and/or\ud imbalance, can be converted to an equivalent encastre case. This allows for the theoretical possibility of eliminating the\ud first simply supported critical speed, thereby reducing the need for very large lateral critical speed margins, as this\ud requirement constrains design flexibility. Although the analysis is performed on a sub 15 MW gas turbine, it is anticipated\ud that this mechanism would be beneficial on any shaft system with high-flexibility/shaft deflection

Deep Learning has attracted considerable attention across multiple application domains, including computer vision, signal processing and natural language processing. Although quite a few single node deep learning frameworks exist, such as tensorflow, pytorch and keras, we still lack a complete processing structure that can accommodate large scale data processing, version control, and deployment, all while staying agnostic of any specific single node framework. To bridge this gap, this paper proposes a new, higher level framework, i.e. Nemesyst, which uses databases along with model sequentialisation to allow processes to be fed unique and transformed data at the point of need. This facilitates near real-time application and makes models available for further training or use at any node that has access to the database simultaneously. Nemesyst is well suited as an application framework for internet of things aggregated control systems, deploying deep learning techniques to optimise individual machines in massive networks. To demonstrate this framework, we adopted a case study in a novel domain; deploying deep learning to optimise the high speed control of electrical power consumed by a massive internet of things network of retail refrigeration systems in proportion to load available on the UK National Grid (a demand side response). The case study demonstrated for the first time in such a setting how deep learning models, such as Recurrent Neural Networks (vanilla and Long-Short-Term Memory) and Generative Adversarial Networks paired with Nemesyst, achieve compelling performance, whilst still being malleable to future adjustments as both the data and requirements inevitably change over time.

No abstract