Smart Grid

October 09, 2012

Smarter Grid Solutions Gets 'Embedded' with McObject Database System

Real-time and embedded software is increasingly being used by electric utilities to manage growing volumes of complex data in support of advanced application features. Now, London-based Smarter Grid Solutions, the UK’s largest dedicated smart grid technology company, has integrated a distributed in-memory database system from McObject (News - Alert) into its Active Network Management technology.

Smarter Grid Solutions' patented and proven technology increases existing grid capacity; facilitates timely connection of new loads and generators; and manages complex, bi-directional, distributed power flows, while maintaining the power system within statutory voltage limits. Its solutions help power companies to accelerate the move toward large-scale use of renewable energy at a fraction of the cost, simply by making better use of the existing grid and simplifying connection for new low-carbon technologies.

Providing power grids with the intelligence to accomplish these goals requires a lot of critical field data —and that’s where eXtremeDB Cluster comes in, according to Neil McNeill, head of Solutions at Smarter Grid Solutions. McObject’s database system resides in Smarter Grid Solutions’ platform software, which monitors and controls devices across the network and facilitates communication between them. The platform runs autonomously in locations ranging from regional control centers to local electrical substations, executing deterministic algorithms to manage multiple generators associated with different network constraints. “Speed of response is critical,” McNeill said. The eXtremeDB Cluster’s time-sensitive roles include tracking conditions at network constraint locations (“pinch points”); and triggering remedial actions on an escalating basis—such as curtailing the energy production from renewable generation or tripping circuit breakers in order to maintain the electrical power system within statutory operating limits, he explained.

Image via Shutterstock

To provide the needed speed, eXtremeDB Cluster relies on its core in-memory database system (IMDS) design. In-memory databases store records in main memory, eliminating the overhead of disk and file I/O, cache management and other functions found in traditional on-disk database management systems (DBMSs). The Smarter Grid Solutions technology’s responsiveness also benefits from eXtremeDB Cluster’s in-process architecture: The database system is embedded in the application process, eliminating the latency of inter-process communication (IPC (News - Alert)) between client and server modules.

Smarter Grid Solutions chose Java for development because of its wide acceptance by utilities and other power network operators. eXtremeDB’s Java Native Interface (JNI) delivers fast performance in this environment by providing direct access, in Java, to database functions that execute as ultra-fast compiled C/C++ code. It also streamlines coding by enabling developers to access eXtremeDB, while working entirely with "plain old Java objects" (POJOs). Smarter Grid Solutions also makes use of Real-Time Java to retain determinism within its Smart Grid algorithms.

In its need for determinism, McNeill compared his company’s technology incorporating eXtremeDB to an automated control unit within a real-time process control system. It must respond synchronously to events on the network, with a time threshold of approximately 800 milliseconds (ms). However, Smarter Grid Solutions has tested the system successfully down to 20 ms, he said. (This overall system response includes database processing with eXtremeDB.) 

Smarter Grid Solutions must meet power network operators’ stringent fault-tolerance requirements. eXtremeDB’s clustering capability ensures database up-time by spreading processing across multiple hardware nodes, providing continuous availability in the event of failure on one node. The Cluster’s “shared nothing” architecture eliminates reliance on a shared storage area network (SAN) or other storage resource.

Clustering also can dramatically increase net processing power available for data management, and reduce cost-of-ownership by enabling the system to expand by adding low-cost servers. This “horizontal scalability” meets Smarter Grid Solutions’ need for a technology that supports both small and large deployments, McNeill said.

“Energy production and transmission solutions, with their need for the highest performance and reliability, present significant challenges to database systems. McObject has long targeted this market with eXtremeDB and can point to many successful deployments,” said Chris Mureen, who is Issaquah, Washington-based McObject’s chief operating officer. “

Edited by Allison Boccamazzo