**A Distributed Computing Architecture for the Large-Scale Integration of Renewable Energy and Distributed Resources in Smart Grids**

Ignacio Aravena, Anthony Papavasiliou and Alex Papalexopoulos

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/67791

#### Abstract

[32] D. Kempe, J. Kleinberg and E. Tardos, "influential nodes in a diffusion model for social networks," in Proceedings of the 32nd International conference on Automata, Languages

[33] D. Kempe, J. Kleinberg and E. Tardos, "Maximizing the spread of influence through a social network," in Proceedings of the ninth ACM SIGKDD international conference on

[34] Y. Hock-Yeow and L. Tong-Ming, "Trusted Social Node: Evaluating the Effect of Trust and Trust Variance to Maximize Social Influence in a Multilevel Social Node Influential

[35] J. Novak, "The Theory Underlying Concept Maps and How to Construct and Use Them,"

[36] J. Novak, "Learning, creating, and using knowledge: concept maps as facilitative tools in schools and corporations," Journal of e-Learning and Knowledge Society, vol. 6, no. 3, pp.

[37] C. Lin and R. Nayak, "A case study of failure mode analysis with text mining methods," in 2nd International Workshop on Integrating Artificial Intelligence and Data Mining,

Diffusion Model," in Springer Lecture Notes in Computer Science, Beijing, 2016.

and Programming, Berlin, 2005.

20 Recent Progress in Parallel and Distributed Computing

21–30, 2010.

Australia, 2007.

Knowledge discovery and data mining, New York, 2003.

Institute for Human and Machine Cognition, United States, 2006.

We present a distributed computing architecture for smart grid management, composed of two applications at two different levels of the grid. At the high voltage level, we optimize operations using a stochastic unit commitment (SUC) model with hybrid time resolution. The SUC problem is solved with an asynchronous distributed subgradient method, for which we propose stepsize scaling and fast initialization techniques. The asynchronous algorithm is implemented in a high-performance computing cluster and benchmarked against a deterministic unit commitment model with exogenous reserve targets in an industrial scale test case of the Central Western European system (679 buses, 1037 lines, and 656 generators). At the distribution network level, we manage demand response from small clients through distributed stochastic control, which enables harnessing residential demand response while respecting the desire of consumers for control, privacy, and simplicity. The distributed stochastic control scheme is successfully tested on a test case with 10,000 controllable devices. Both applications demonstrate the potential for efficiently managing flexible resources in smart grids and for systematically coping with the uncertainty and variability introduced by renewable energy.

Keywords: smart grids, stochastic programming, asynchronous distributed algorithm, stochastic control, demand response

### 1. Introduction

The progressive integration of renewable energy resources, demand response, energy storage, electric vehicles, and other distributed resources in electric power grids that has been taking place worldwide in recent years is transforming power systems and resulting in numerous

© 2017 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

operational challenges, including uncertainty of supply availability, distributed storage management, real-time coordination of distributed energy resources, and changing directions of flow in distribution networks. These challenges demand a shift of the traditional centralized power system operations paradigm toward the smart grid paradigm [1], where distributed computing and control stand out as a promising technology with the potential of achieving operations with optimal performance.

The academic literature includes various applications of distributed computing in power system operations, including long- and mid-term planning, short-term scheduling, state estimation and monitoring, real-time control, and simulation [2–5]. Early studies pointed out several challenges related to communications and the heterogeneous characteristics of distributed computing systems, which needed to be addressed first in order to implement distributed computing applications. Nowadays, standard communication protocols are a mature technology and most current distributed computing resources can perform a broad range of operations. Such advances in distributed computing technology have paved the way for developing and implementing scalable distributed algorithms for power systems.

The prevailing industry view, as we move forward into the future smart grid, is that it will entail: (i) broadcasting of dynamic prices or other information and (ii) telemetry backhaul to market participants. In the proposed model, distributed energy resource aggregators are often regarded as transaction brokers between end customers and various upstream market participants. The "failure-free market" design for a pure market-driven solution under this paradigm has been elusive, despite decades of research and development. In this chapter, we analyze the deployment of distributed computing as an enabling tool for managing the shortterm operations of smart grids in two levels:


Centralized clearing at the high-voltage grid level and distributed clearing at the distribution grid level can be integrated in a cooptimization framework, as recently proposed by Caramanis et al. [10]. These two applications of distributed computing in power system operations demonstrate the potential to fully harness the flexibility of the grid and smoothly integrate large shares of renewable and other distributed energy resources in power systems without deteriorating the quality of service delivered to consumers.

The rest of the chapter is organized as follows: Section 2 introduces the deterministic and stochastic unit commitment problems. Section 3 proposes an asynchronous algorithm for solving SUC and presents numerical experiments on a network of realistic scale. Section 4 presents the ColorPower architecture for managing demand response in the distribution grid and demonstrates its capability through a numerical experiment. Finally, Section 5 concludes the chapter.
