Increasing network capabilities for renewables with Big Data

As transmission and distribution systems become increasingly challenged by rapid changes in the generation portfolio, modern power electronics and big data analytics are helping network operators to get far more from existing assets. As a result, more renewable energy can be added to the grid without compromising reliability.

A major underlying trend affecting transmission and distribution systems is the transition from large central and typically thermal baseload and peaking power plants to the variable output renewable sources of generation that are widely distributed across the network. Rapidly growing sources of electricity generation like wind and solar are presenting challenges for T&D system operators (TSOs and DSOs) because the network is now required to operate in ways it was never designed for.

For example, in Germany, power is flowing from the north and eastern coastal regions, where large volumes of wind capacity are installed, down to the major cities and load centers in the south of the country. The loading of existing assets is changing but, perhaps more significantly, the situation is changing very rapidly. A new PV plant can be built in a matter of months for instance. Big Data and the analytical capabilities that it can support is emerging as a key tool to address this growing challenge.

Increasing the performance of existing assets
Potentially, two issues are of primary concern for grid owners and operators. Maintaining grid stability in terms of voltage and frequency control is an obvious requirement, but it is also crucial to maintain reliability as parts of the network are being loaded and taxed far more than had previously been required. Thus, maintaining the grid within desired and often mandated performance parameters means managing two factors – physical infrastructure as well as the dynamics of the grid in operation.

In the past, the transmission and distribution network operators would invest in reinforcing the network by building new lines and substations. Now though, in many regions, grid strengthening is taking a lot longer than the development of renewables. Building a new line may take 10 or 20 years in applying for all the consenting paths land acquisition and so on. As a result, demands on the network are outstripping the traditional capabilities of the TSOs and DSOs to reinforce and extend its capabilities.Under these circumstances, TSOs and DSOs have to make the grid stable without using a lot of new power lines whilst integrating additional geographically distributed renewables on the network.

One mechanism that can help achieve this goal is to operate the existing network closer to the edge of stability, than we have previously been accustomed to. This is only possible by knowing the precise condition of the network. By applying additional sensors and monitoring equipment, it is possible to determine more exactly than before what situation the network is operating in, as well as the status of each of the network assets.

Gathering this data and using sophisticated analytics allows operators to more accurately predict if one of the assets might fail in the future. Whereas, the previous generation of grid management might operate with a 30 per cent or greater operational safety margin; more granular data coupled with precise analytics enables operators the scope to narrow this margin and recover more performance from existing assets. By reducing the operational ‘safety’ margin to perhaps 10-15 per cent, far more power can be delivered through existing assets without increasing the risk of failures.

For example, a major power plant may have two lines connecting it to certain loads. Networks,today, are designed so that evenin the event of a failure either power line can handle the full load alone. This (n-1) criterion means that if one component in the grid fails, the power supply must not be impacted at all. Under such a scenario, this network is utilised at say 30-40 per cent of maximum in all but the rarest of circumstances. But, knowing exactly what power is flowing in real time allows operators to utilise the existing network much more. Thus, by predicting a particular line might fail in the next, say, half a year operators can execute some remedial maintenance action to ensure that the network is still operating within limits, although without utilising the redundancy of the (n-1) security criterion that had previously been a requirement in network planning.Key to this approach is real-time data informing operators of the actual status of network assets.

Using Big Data to expand network capabilities
Executing a strategy of utilising the existing network better demands having critical information on the exact status of each asset, the power flows and the broader situation of the network. The second element is the requirement for increased capabilities within the power network to use this information to respond to fluctuations. Siemens, for example, has developed its STATCOM device, SVC PLUS, to deliver precise amounts of inductive or capacitive reactive power as required in order to maintain voltage stability. STATCOM also delivers large volumes of data on the network conditions and its own performance.

Based on Modular Multilevel Converter (MMC) architecture, STATCOM uses Voltage-Sourced Converters (VSCs) to provide a near perfect leading or lagging sinusoidal waveform independent of the AC system voltage. Able to respond reflexively or rapidly interact with the existing power flows in a matter of miliseconds, STATCOM is adaptable and provides longevity. It also employs relatively few robust and proven components, such as typical AC power transformers, reactors, capacitors and industrial Insulated Gate Bipolar Transistors (IGBTs).

The data STATCOM supplies can be used to build an accurate relationship model of the network making it possible to digitally ‘tune’ the network. For example, a good relationship model of the network can reveal that in the next 30 minutes the power flowing across a heavily loaded line will increase even more. Under these conditions operators must take steps to reduce loading on that asset to build a sufficient performance margin.

Analytical tools combined with very accurate simulation models such as a digital twin of the network and other measures can help with network planning and support the TSOs and DSOs in keeping their systems stable.

Securing data through the cloud
One recurring question faced by those employing big data concerns cyber security and data handling. Siemens has developed a secure cloud-based data platform for IoT called MindSphere. The MindSphere platform is used to push all the data generated by devices like STATCOM to central cloud storage to support asset management, additional forecasting and stabilisation of the network.

It organises the data to allow users to create applications which then deliver appropriate additional information. For example, applications may identify which STATCOM units are used most often or which are delivering reactive power at maximum performance and how often. It allows other market players to develop appropriate applications and analytical tools. Adding further STATCOM units can increase these kinds of capabilities even more for the smart asset management for the operation or analysis of the grid.

The heart of Big Data functionality for transmission and distribution networks is the secure Mindsphere platform where data is gathered in a single space. It will be crucial to include all portfolio assets as well as additional centers to have one combined set of data which can be accessed and analysed.

Extending smart network capabilities
Introducing more power electronics like STATCOM to the network is adding far more capability and flexibility to the transmission system. As power electronics become cheaper and smaller, this capability is expanding downstream into the distribution networks too. This is an increasingly important consideration for DSOs which are also facing the challenges associated with increasing renewables penetration as well as emerging developments like battery energy storage and electric vehicles.

Solutions like MVDC (medium voltage DC transmission) but also a STATCOM development to lower power ratings will see many of the problems that are currently being addressed at transmission level solved also within the distribution networks.

Looking forward, it is possible that installing similar equipment across the distribution network can reduce loads and reduce the loading on the transmission side too. Thus,Big Data analytics might help to reduce the requirement for investment in assets and reinforcements of the whole grid. However, in many jurisdictions, currently, the TSOs and DSOs are separate entities and with multiple stakeholders market structures, it may be an obstacle to such a development. Nonetheless, optimisation of each regional distribution system would be beneficial for the transmission system and reduce costs for the complete electricity supply network.

Furthermore, the capabilities of technologies like STATCOM are such that as the grid code evolves to reflect changing conditions for the TSOs and DSOs, investment in the maximum flexibility for the network supports future changes to the generation portfolio with the inclusion of, say, storage or more renewables but also potential changes to the grid code.By investing in maximum flexibility, the TSOs and DSOs can future proof their networks for an uncertain future, releasing the full capabilities of their network assets and in turn allowing more renewable energy generation capacity to connect.

Authored by:

Dr. Holger Mueller
Head of Strategy and Principal Expert
Energy Management Division
Siemens Transmission Solutions