New Challenge for Automation
Concentrating solar power plants are called upon to perform complex operations in harsh environments precisely and automatically.
Concentrating solar power (CSP) plants represent a new challenge for automation. The ratio of about 400 I/O signals per MW is one of the highest in power generation. The system tags also have an extremely high ratio of about 3500 tags per MW. A parabolic trough CSP plant is one of the most automated types of power plant in the energy-generation industry, comparable to a nuclear power plant.
In addition, the overall architecture is highly distributed, covering thousands of square meters in the solar field, a factor that increases the complexity of plant communications.
The automation of the balance of a plant is based mainly on maintaining the working fluids (oil, heat-transfer fluid, molten salt or water/steam) within narrow operating margins of temperature and pressure control, as well as on controlling unstable steam conditions due to the nature of the sun’s irradiation.
CSP internal bus communications are also a challenge. About 150,000 tags are communicated through the system buses, which are typically Modbus TCP or Profibus DP. Communication from the solar field to the central power plant database has to provide a complete solar field update in less than five seconds.
The distributed control system (DCS) has to be highly reliable (with 99.99% availability). For this reason, redundancy has to be provided for the control processors, the I/O cards for all analog I/Os for critical control loops, the communication and interface processors, the data storage devices, communication networks and power supplies—so that single component failures will have minimal effect.
The software communication system (SCS) is responsible for managing the huge volume of data traffic and for controlling some of the solar field functions related to the overall behavior of the plant—such as the emergency modes, start-up modes, etc.—and it is fully integrated with the plant (DCS) hardware. Depending on customer specifications, the SCS can be equipped with the same hardware processor as the DCS or with an industrial PC. This subsystem is critical for securing any kind of point-to-point, event-driven or broadcast communication.
Local Automation Controllers
In the case of CSP technologies, the solar field of each solar thermal plant consists of a system of long pipes in which the fluid is heated by the reflection of the sun’s rays and concentrated by an optical mirror system onto the pipes. The average number of local automation controllers (LOCs) per 50 MW is around 450–640 units, depending on the solar field design and whether the plant has thermal storage. These several hundred devices are connected to a communication bus system that is managed by the SCS in order to avoid communication collapse and to secure data integrity and the time cycle.
The LOCs should accurately calculate the expected position of the sun, as per the National Renewable Energy Laboratory (NREL) model, and drive the mirror system to track the sun for closed-loop control.
Control of the solar field and its coordination with the balance of plant processes and electrical systems is provided by the LOC, in order to ensure that the temperature and pressure are optimal for each moment of the day and that safety conditions like high winds, hail and overpressure conditions are taken into account.
LOC Design Principles
The control part of the LOC is performed by proven standard industrial microprocessor-based programmable logic controllers (PLCs).
The LOC should have a modular design to facilitate easy system adaptation to future requirements and achieve an operating life of at least 20 years. The system should have the capability and facility for expansion through the addition of controller modules, I/O communication cards, etc., as well as for future software modifications.
The LOC should be easy to maintain. Each hardware part of the LOC should be supplied by an industrial standards manufacturer and be able to be replaced within a reasonable time anywhere in the world.
Each component and system part should be of proven reliability. The minimum reliability of each item of equipment—such as the electronic modules/cards, power supply and peripherals—should be such that the availability of the complete system is assured for 99.7% or higher.
The design of the LOC and related equipment should adhere to the principle of fail-safe operation wherever the safety of personnel and plant equipment is involved. Fail-safe operation means that the loss of a signal or the loss of excitation or failure of any component should not cause a hazardous condition. It also means that the occurrence of false trips is avoided.
As pointed out earlier, the control part of the LOC should be performed by proven standard industrial microprocessor-based PLCs, not by a low-cost dedicated electronic board that is likely to create operational problems for maintenance.
The LOC is responsible for detecting and controlling the angular position of the mirror system by means of sensors based on an incremental encoder or inclinometer.
The LOC PLC is provided with the type of memory and capacity to store data and parameters permanently. The angular position of the collector should be stored in a permanent memory, even if an electrical blackout occurs.
LOC internal software must be able to calculate the expected position of the sun and the corresponding solar vector to which the mirror must be pointed.
It should calculate the theoretical position of the sun, not just “see” the position of the sun. The reason for this is to avoid misreading the sun’s position due to the presence of clouds or other disturbances. These calculations should follow the NREL system and have a control algorithm of ±0.0003º, which is extremely demanding for an industrial process.
The software architecture should be flexible to permit future modifications, so that dedicated hardware solutions will not be a problem for the maintainability of the overall system.
The LOC communicates with the plant DCS and must be able to deliver the following information to the DCS:
- Mirror identification
- Date and time
- Configuration parameters
- Calculated solar vector
- Alarm status
- Operation mode
- Mirror position set-point angle
- Fluid temperature
- Internal temperature of the LOC.
Extreme Environmental Conditions
The real challenges for the local automation of the solar field are heat, ultraviolet radiation and dust.
It is easy to imagine what the outdoor LOC equipment in a CSP plant in a desert location is subjected to. The heat from the sun’s irradiation is intense and equipment exposed to ultraviolet rays ages quickly. The electronics have to live with their worst enemies—the heat and dust. And, because customers require completely sealed solutions, ventilation holes and movable parts like fans are not permitted.
That is why the ABB Group (Zurich, Switzerland) has designed a cubicle enclosure with double walls to withstand desert conditions and direct ultraviolet radiation. The external enclosure of the LOC is IP65/ NEMA 4X plastic molded. It is attached to the metallic structure of the drive pylon (the central pylon of the collector) without any protective shadow.
The enclosures are designed for 25 years’ durability. The mechanical design can withstand extreme temperatures from -5ºC to 55ºC and relative humidity of 25–95%, including condensation and corrosive vapors.
For each CSP project ABB engineers design the optimal allocation of components in the LOC to ensure that the thermal behavior of the heat generated in the LOC does not affect the functionality, taking into account that the heat has to be dissipated only by means of passive solutions, without holes or fans in the cubicle.
Each LOC prototype is tested in the expected operating conditions and with the dimensions for the maximum ambient temperature of the site (typically 40–45ºC) and 1000 W/m² solar radiation, and with internal thermal dissipation. The internal temperature has to be at least 5ºC below the least heat-resistant component. Every design is verified and certified in an independent laboratory.
The LOC is provided with its own Ethernet communication port to the DCS by way of TCP/IP Modbus. Multimode fiber optic or copper RJ45 links connect each LOC to the next LOC within each network ring.
Communication with any LOC does not inhibit, delay or disturb communication with other LOCs, the DCS or other devices in the network. The configuration program and configuration updates for each LOC are downloaded over the communication network.
A serial communication port is provided for local configuration and troubleshooting via the Ethernet communication port. For maintenance purposes, the LOC system includes a software tool to link a PC or handheld device with any LOC in the solar field from any point in the communication network. It includes control and communication software and human interface screens for any LOC or group of LOCs.
The LOC system has accurate time synchronization with the plant DCS. Real-time clocks in the control equipment are synchronized to within 1 ms by a GPS clock installed at the plant. A network time protocol (NTP) server synchronizes the clocks of all on-site LOC systems at regular intervals.
ABB’s 1-GW Installed Base
ABB has many years’ experience in automating CSP solar fields in Spain, the US and Egypt, as well as active involvement in many CSP and ISCC (integrated solar and combined cycle) power plant projects in Morocco, South Africa, Australia, China, India, Chile, and other countries. The company’s large installed base in these projects now amounts to more than 20 installations with a combined generating capacity in excess of 1 GW.
This article first appeared in ABB In Control, a publication of ABB Power Generation, a provider of integrated power and automation solutions for conventional and renewable-based power generation plants and water applications. ABB Power Generation is part of the ABB Group (Zurich, Switzerland). SME is grateful to ABB for granting permission to use this article.
Published Date : 11/11/2013