Case Study: Kiira Power Station

Introduction

Distributed control systems in industrial processes are intended to provide communication between distributed controllers to achieve efficiency and reliability of the plant. Power plants especially hydro power plants have a design life of at least 50 years. Unfortunately, their control systems have a much less life of between 10 to 15 years. As such, the plants begin to experience faults associated to aged controls, and this is no exception for the distributed control systems (DCS). An old or legacy DCS can be defined as one that is difficult or impossible to modify and or evolve to the latest state of the art technology in terms of both software and hardware. As the DCS ages, they experience several defects and shortcomings such as CITATION Jes06 \l 1033 [1]:

Costly and time-consuming fault tracing and software maintenance due to missing documents as well as lack of knowledge of the system software and internal functionality.

The obsolete software and hardware on which the systems run is costly to maintain and usually very slow.

Expansion of old DCS is either impossible or very expensive

Lack of clean interfaces makes integration difficult

As technology evolves and the need for power grows, power plants will require migrating legacy DCS to newer environments allowing for easy maintenance and adaptation to demand while maintaining functionality of the entire system without doing a complete re-development of the system. This justifies the need to look at how and when to migrate an existing DCS.

Case: Kiira Power Station

Kiira Power Station is a hydro power plant with a net installed capacity of 200MW derived from five vertical shafts; fixed blade Kaplan turbine machines each generating 40MW at a head of 21m. The first and last units were commissioned in 2000 and 2007 respectively. The power plant uses an ABB distributed control system of the ADVANT family, made of two operator stations-HMIs (which are used to start, regulate, monitor parameters and shut down units), five-unit computers, one station computer, one switch yard computer, of the type AC410 while the one auxiliaries computer is an AC70. The OEM declared both AC410 and AC70 obsolete in 2016 and 2015 respectively.

The Station Computer and switchyard computer are non-operational because of hardware problems that cannot be fixed due to unavailability of spares. According to the manufacturers product life cycle management schedule, this model of the ADVANT (AC 410) has been declared obsolete. The switchyard computer provides a communication interface between the 132kV switchyard in NPS, KPS generating units and Lugogo Control Centre, which monitors the countrys power generation and distribution. The KPS station computer relays signals from various transducers and equipment around KPS to the KPS process computers, significantly boosting plant monitoring and diagnosis in case of a failure.

It is thus necessary for the DCS controllers to be replaced with new versions of technology that are still being manufactured, supported, have spares on market. In such a case as KPS, the engineers need to consider many factors when planning the upgrade project, such as, replacing the entire network at once, or doing it in phases. They also need to consider whether to maintain the existing OEM or change to another one. Justification

Hydro power plants are designed to last for over 50 years of service. To maintain efficiency and reliability of the plants, many rehabilitations and refurbishments are done over the years. As such, the control systems of such long serving plants also require upgrades and migrations as they soon reach the end of their lifecycle, becoming obsolete. Towards the end of the DCS lifecycle, human resource knowledgeable about the system become scarce due to retirement and product spares and technical support become scarce. As the control system ages, sudden malfunctioning of the system is inevitable and this usually results into sudden process shutdowns, long downtime due to lack of spares and support accelerated by incompatibility of the current market technology with the old/legacy systemsCITATION Upg \l 1033 [2].

As technology changes, there is rapid development in the field of process and industrial automation forcing the vendors of DCS to upgrade their systems to accommodate the latest-state-of-the-art technology. With such advancements, production lines for legacy systems are terminated, phasing out and reducing technical and spares support for the legacy system. The vendors then, advise users of the legacy systems to upgrade to the latest systems to avoid the hard-hitting impact of obsoleteness of equipment and sudden component failure. It is thus important to plan for the power plant control system taking into consideration what strategy to use to maintain and upgrade old/legacy control systems with as minimum interruption to production as possible.

Thesis objectives

This research aims at analyzing upgrading and migration strategies for legacy/old distributed control systems to the state-of-the-art control systems, minimizing cost, risk and process downtime.

The specific objectives towards addressing the issues around migrating a DCS include:

Avail guidelines to establish when a DCS is due for replacement: this will be achieved by an analysis of the projects undertaken by some of the DCS vendors on the market shall be done to establish what motivated their clients to upgrade their DCS.

Review common factors that motivate for a DCS upgrade or migration: A case study on old hydro power plants with old control systems will be done to establish the failure rates and trends of the old control systems. For those with plans of upgrading, the motivating factor to this decision will be assessed.

Analyze strategies that can be followed for existing DCS upgrade: From the vendor projects, the approaches used to upgrade or migrate the clients DCS will be examined taking into consideration the challenged and advantages attained.

Provide general guidelines and methodologies on how to upgrade a DCS: Is it feasible to mix vendor products on the same DCS? An analysis of power plant using mixed vendor products or third-party production their control networks will be studied.

Thesis overview

This thesis therefore attempts to address the fore mentioned concerns as per the chapters below:

Chapter 2: The literature review analyses a typical DCS, looking at the architecture of a typical DCS for a hydropower plant with Kiira power station as the case study. It will also entail related work in terms of software and hardware DCS upgrade.

Chapter 3: it looks at the case studies of DCS upgrades for at least three DCS vendors on the market today. Details of motivating factors for the upgrades are analysed, common failure trends of the systems as well as upgrade strategies used for the different case studies.

Chapter 4 covers KPS DCS in detail, highlighting the failures and upgrade solutions or strategies to consider. Analysis of research findings is done and future works as well as recommendations to consider are stated in chapters 5 and 6.

Chapter 2: Background

Hydro power plant overviewAs seen from figure 1, a hydro power plant of any size and complexity is basically made of turbines, generators, breakers, transformers and subsystems installed with numerous field instruments, valves, operator interfaces and other devices working together to make the plant run efficiently and reliably. All those parts are connected and controlled by a DCS that is considered a collection of processes/controllers and input/output systems capable of making everything work together. The control logic of the controller related to each generating unit includes: start/stop sequence, units circuit breaker handling, active and reactive power control, temperature control, communication with protection system, energy calculation and annunciation system.

Figure SEQ Figure \* ARABIC 1: Typical power plant layoutCITATION ABB00 \l 1033 [3]Distributed Control Systems (DCS)A DCS is a control system designed to control field devices and process computers that are geographically distributed across the process plant or industry. A very fast communication network is used to connect the operator computers and field devices to these controllers. An example of field devices on the network are sensors and actuators that are connected to the output and input modules of the controllers. Figure 1 shows a basic layout of a DCS network that is made of a local area network, general-purpose computer, data highway link, local display, local control unit, field devices and a data acquisition unit.

Figure 2: DCS Network layout CITATION Kin02 \l 1033 [4]

The DCS enables control of the generating unit from the operator station or control console in control room and/or from local HMI or manual controls on the unit control cubicles. The DCS is thus designed to ensure that supervision and control of all major power system equipment can be done from control room (remotely) ensuring voltage and frequency stability, coordination of generators, optimizing and fault finding tools, and allowing for local viewers support.

Critical as the DCS may be, most power plants have their controllers as old as their civil structures, that is, their DCS platforms may be over 25 years of age. In terms of computing power, these systems are extremely slow. Studies from different vendors show that such power plants have updated parts of their networks except for the controllers/processors and most of the input and output hardware is ancient. Figure 2 shows are comparison in terms of the civil structure and DCS component life cycle.

Figure 3: Life cycle compari...