Proc. XXV International Scientific Conference Electronics - ET2016, September 12 - 14, 2016, Sozopol, Bulgaria

978-1-5090-2881-8 /16/$31.00 ©2016 IEEE

Experimental Study of Cloud Computing Based SCADA in Electrical Power Systems

Dimitar Grozev, Grisha Spasov, Mitko Shopov, Nikolay Kakanakov, Galidiya Petrova Department of Computer Systems and Technologies, Department of Electronics, Faculty of Electronics and Automation

Technical University of Sofia, branch Plovdiv 25 Tsanko Diustabanov Str., 4000 Plovdiv, Bulgaria

dimitargr@gmail.com, {gvs, mshopov, kakanak, gip}@tu-plovdiv.bg

Abstract – The paper presents a test bed platform for implementation of Cloud Computing in SCADA systems for Electrical Power Systems. The general idea of the experimental study is to complete operation tests of traditional SCADA systems on cloud computing platforms, and integrate the main application modules of SCADA systems for electrical power metering on the cloud platforms. The logical networking structure of the Control Center of SCADA systems is presented. The virtualization of the different type of servers in the Control Center together with physical networking solution of the cloud platform realization is described.

Keywords – Cloud computing, SCADA, Power systems

I. INTRODUCTION An electric power system typically involves generation, transmission, transformation, distribution, consumption, and dispatching processes. Electric power is generated and consumed simultaneously and is not available for mass storage or transportation. Thus, as the key to the electric power system, monitoring and dispatching guarantee the reliability and security of electric power generation, transmission, distribution, and consumption, and play a key role in providing high quality and economic electric power. An electric power dispatching and monitoring system is often called a supervisory control and data acquisition (SCADA) system, energy management system (EMS), or distribution management system (DMS) globally. In this article, we call it the electric power SCADA system. In the early stage, the electric power SCADA system mainly used multi computer architecture, consisting of single server and two server cluster systems. Currently, the system uses computer systems with a distributed open architecture [1]. The electric power SCADA system provides SCADA, automatic generation control (AGC), automatic voltage control (AVC), EMS, DMS, dispatcher training system (DTS), geographic information system (GIS), and other useful functions. The software application system is constructed and developed by using the standard CIM model, and software architecture has evolved from the Client/Server architecture to the current Browser/Server architecture. The new generation Smart Grid [2] dispatching and control system uses multi-core computer cluster technology to improve system reliability and processing capacity, and uses a service oriented architecture (SOA) to enhance system interoperability and achieve horizontal integration and vertical interconnection

of power grid dispatching services. Hardware infrastructure of these application systems involves high performance servers, complex high speed computer networks, high performance and highly reliable data storage systems, and workstations. The SCADA software is developed based on the platform which consists of Windows, UNIX, and Linux operating systems, and is based on relational databases. The whole system is connected through computer networks for data exchange and sharing, and application programs share information through an enterprise service bus (ESB). With the development of the traditional IT application systems, IT-based applications have been expanding deeply to another industry field, and encountered various problems and bottlenecks. The same is true for the electric power SCADA system, which is a professional IT application system. The future electric power dispatching center should have high computing capacity, and powerful information acquisition, integration, and analysis functions especially with the advancement of Smart Grid applications. Existing centralized computing platforms of electric power systems can hardly meet the above requirements, which have become one of the major bottlenecks in the Smart Grid. Some of the major disadvantages of these platforms are as follows [3]:

• Low basic resource utilization and poor scalability. A large amount of basic computing resources to meet the demands in peak hours are idle during off peak hours. To ensure reliability, lots of resources are redundant, and cannot be fully utilized. Contradictions are growing between energy demands and conservation policies. Due to the upgrade of business, the existing IT infrastructure cannot be reused. Currently, analysis and computing in the electric power system rely on the centralized computing platform in the dispatching center. Due to limited computing capacity, poor scalability, and high upgrade costs, large scale power systems suffer from insufficient data storage and analysis capabilities.

• Poor system interoperability leading to information islands. Parallel application systems have their own architectural features, and therefore resources cannot be exchanged or reused.

• Increasing management costs and risks, and decreasing equipment utilization.

96

II. SCADA SYSTEM COMPONENTS The traditional components of SCADA systems are shown on figure 1.

Fig. 1. SCADA network architecture

1. RTU – remote terminal units are connected to

sensors. They have telemetry hardware capable of sending digital data to the supervisory system, as well as receiving digital commands from the supervisory system.

2. PLCs – programmable logic controllers are also connected to sensors. PLCs have more sophisticated embedded control capabilities than RTUs. PLCs do not have telemetry hardware, although this functionality is typically installed alongside them. PLCs are sometimes used in place of RTUs as field devices because they are more economical, versatile, flexible, and configurable

3. Telemetry system is used to connect RTUs and PLCs to control center or data warehouse systems.

4. A data acquisition server is a software service which uses industrial protocols to connect software services, via telemetry, with field devices such as RTUs and PLCs. It allows clients to access data from these field devices using standard protocols.

5. HMI is the apparatus or device which presents processed data to a human operator, and through this, the human operator monitors and interacts with the process.6. A Historian is a software service which accumulates time-stamped data in a database which can be queried or used to populate graphic trends in the HMI. The historian is a client that requests data from a data acquisition server.

6. A Historian is a software service which accumulates time-stamped data in a database which can be queried or used to populate graphic trends in the HMI. The historian is a client that requests data from a data acquisition server.

7. A supervisory (computer) system, gathering (acquiring) data on the process and sending commands (control) to the SCADA system.

8. Communication infrastructure.

III. CLOUD COMPUTING

TECHNOLOGIES FOR ELECTRIC POWER

INDUSTRY Cloud computing can be used to solve various problems occurred during application deployment, use, and innovation process of the electric power SCADA system. Nowadays most of companies have their own cloud platform and virtual infrastructure, where they run business critical application and store data.

Private cloud advantages [4]:

• It is reliable and scalable. All resources are virtualized and in case of demand more storage or computing without downtime or impact can be added.

• Fast provisioning. Using techniques like templates can deploy thousands machines with few clicks.

• Automation. Pretty much everything can be automated using powercli or REST API. Common user interface: decoupling the computation infrastructure and the input system, enables multiple user interfaces to exist side by side allowing user-centric customization.

• Cost effective.

IV. PRIVATE CLOUD

INFRASTRUCTURE

For the purposes of proposed experimental study a private cloud infrastructure is used. It is build and running in Technical University of Sofia, branch Plovdiv and its physical topology is shown on figure 2.

Fig. 2. Private cloud physical topology

97

The private cloud in use is shared by the university departments and students. Everyone can request computing capacity and after an approval process one could receive VMs with requested parameters. The following types of traffic are recognized:

• Management – this include all traffic generated by virtual infrastructure: vMotion, iSCSI, NFS, fault tolerance, web-based graphical user interface, etc.

• VM traffic – this is traffic generated from all VMs connected to VLAN based port groups – various types, generated by different applications.

• VXLAN traffic – this include traffic generated from all VMs connected to logical switches - various types, generated by different applications.

• SCADA traffic – this include traffic generated from SCADA system servers

Because of the importance and sensitivity of the SCADA traffic and servers a special logical networking that isolates SCADA traffic is proposed. It is based on VMware NSX [5] technology (figure 3).

Fig. 3. Private cloud network topology

Entry point of SCADA network is scada-dsnet-esr1. This is NSX router with multiple features including firewall, IPSec and SSL VPN support, load-balancing and other features. A set of firewall rules restrict traffic to/from SCADA servers. IPSec tunnels connect scada-dsnet-esr1 to remote RTUs and PLCs. Through this secure channel data for controlled processes are collected. The scada-dsnet-esr1 runs OSFP with scada-dsnet-dlr1. The scada-dsnet-dlr1 is a distributed logical router (DLR) [6] that optimize routing

traffic between different SCADA components. Distributed FW [7] is used to secure the communication.

V. CONCLUSION

The paper presents an implementation of a cloud-based SCADA system for use in electrical power control systems. The private cloud infrastructure in use is based on VMware platform. Deploying SCADA in cloud infrastructure automatically inherits all cloud advantages:

1. Most of SCADA components are virtual machines: DB server, Application Server, OPC server, Engineering and HMI VMs. This means that we can easily create new ones, clone, backup, take snapshots, HW upgrade - change parameters like CPUs, RAM, HDD, software upgrades.

2. All SCADA components can be protected by technologies like HA and FT [7].

3. All communication nodes are virtual nodes and thus are also protected by HA - scada-dsnet-esr1 and scada-dsnet-dlr1.

The presented test bed platform will allow various types of information flows and configurations to be investigated and analysed and the best set-up for the production services to be selected. Based on the experiments with different configuration of network services in the cloud the traffic flows will be optimized for best performance in terms of delay and resource usage.

GLOSSARY vSphere - VMware vSphere is the brand name for VMware's suite of virtualization products.

ESXi/ESX – Elastic Sky X. An enterprise-class, type-1 hypervisor developed by VMware for deploying and serving virtual computers.

VXLAN (Virtual Extensible LAN) – A Virtual Network that emulates an Ethernet broadcast domain.

vMotion – VMware vSphere live migration allows you to move an entire running virtual machine from one physical server to another, without downtime.

NSX - VMware NSX is the network virtualization platform for the Software-Defined Data Center (SDDC).

DRS (Distributed Resource Scheduler) – technology for balancing the computing capacity by cluster to deliver optimized performance for hosts and virtual machines.

VDS – VMware vSphere Distributed Switch (VDS) provides a centralized interface from which you can configure, monitor and administer virtual machine access switching for the entire data center.

DLR – Distributed Logical Router. It separates Control and Data plane. Control plane is a VM, but data plane is part of hypervisor kernel.

98

ACKNOWLEDGMENTS The presented work is supported by the National Science Fund of Bulgaria project “Investigation of methods and tools for application of cloud technologies in the measurement and control in the power system” under contract Е02/12 (http://dsnet.tu-plovdiv.bg/energy/). REFERENCES [1] Arghira N., et al. Modern SCADA philosophy in power system operation – a survey, Scientific Bulletin, University POLITEHNICA Bucharest, Series C: Electrical Engineering, vol. 73, no. 2, pp. 153-166, 2011.

[2] Fang Xi, Misra Satyajayant, Xue Guoliang, and Yang Dejun. Smart Grid – The New and Improved Power Grid: A Survey, IEEE Communications Surveys & Tutorials 2012, vol. 14, no. 4, pp. 944 – 980, 2012.

[3] Gang Tan, Li Zhao, Yaduan Wang, Jingwei Xu, and Honggang Liu. A Safety Design of Electric Cloud Computing Platform, In Proceedings of the Fourth International Conference on Computational and Information Sciences (ICCIS '12). IEEE Computer Society, Washington, DC, USA, pp. 868-871, 2012.

[4] Alamri A, Ansari WS, Hassan MM, Hossain MS, Alelaiwi A, Hossain MA. A survey on sensor-cloud: architecture, applications, and approaches, International Journal of Distributed Sensor Networks, Article ID 917923, Feb 2013.

[5] VMware NSX Technical Product Management Team. VMware® NSX for vSphere Network Virtualization Design Guide ver 3.0, Aug 21, 2014, online [04.02.2016]:

www.vmware.com/files/pdf/products/nsx/vmw-nsxnetwork-virtualization-design-guide.pdf

[6] B. Hedlund. Distributed virtual and physical routing in VMware NSX for vSphere, The Network Virtualization Blog, November 25, 2013. online [04.02.2016]:

https://blogs.vmware.com/networkvirtualization/2013/11/distribut ed-virtual-and-physical-routing-in-vmware-nsx-for-vsphere.html

[7] Debashis Basak, Rohit Toshniwal, Serge Maskalik, and Allwyn Sequeira. Virtualizing networking and security in the cloud. SIGOPS Oper. Syst. Rev. 44, 4 (December 2010), pp. 86-94, 2010.

99