Electrical Systems and Control Methods: Advanced Control Methods
By Ed Butts, PE, CPI
Advanced control systems, such as a sophisticated Supervisory Control and Data Acquisition (SCADA) computerized system with multiple sites, must adhere to a few basic rules for design, installation, and implementation to ensure success.
This includes observing steps such as an inventory and type of inputs and outputs (I/O), sensor values, processing steps, cycle and time delay periods, sequence of operation, a complete detail and outline of normal operating conditions as well as track any abnormal or damaging conditions that could exist or occur during routine operations, along with anomaly and failure modes, redundancy, and protection.
One of the advantages of a well‐designed SCADA system is its ability to continuously record and archive operating data. This is invaluable for reviewing and troubleshooting process issues and verifying regulatory compliance.
This feature also relieves an operator of having to make repeated rounds and manually record process data, later transcribing the information into reports. This feature is commonly called an “historian.”
Depending on the system storage capacity, an historian can store many years of operating data from all water system processes. Modern SCADA systems allow operators to maintain complete awareness and control of their treatment plant and pump station processes; track and modify remote site operation and parameters; record pertinent process data for archiving, troubleshooting, or for regulatory purposes; and easily expand or upgrade process instrumentation without any significant disruptions or outages.
SCADA Systems
Breaking down SCADA systems into two primary categories yields the following:
- Supervisory Control: Devices and systems used to extract and transmit specific parameters or conditions for process and system primary or secondary control and management of a water system (the “SC” in the SCADA acronym)
- Data Acquisition: Primarily intended for use in extraction, collection, and dissemination of setpoints and accumulative data relative to static or operational parameters of a water system (the “DA” in SCADA). The best-known method for creating an understanding of the system requirements comes in the form of a functional specification or narrative. The document carefully lays out in a structured format how the water system’s process is intended to operate and how the control system will interact with that process.
The document may consist of less than a single page describing the control system’s intended operation up to hundreds of pages of detailed programming instructions, ladder diagrams, and hardware specifications.
Any documentation, schematics, system interface drawings, sketches, programming language, etc. that provides a description for where the processes and related equipment are located relative to each other, along with the intended operator monitoring/control stations, should be included.
The need for more sophisticated control, such as analog to digital conversion, cascade loops, or the need to interface to variable frequency drives or other peripheral devices, should be initially identified.
A review of the failure modes and potential outcomes should also be outlined. In many cases, the use of a Failure Mode Effects Analysis (FMEA) is warranted, particularly for complex control systems open to potential daisy-chain events that could shut down or impede a process, individual pump station, or in the worst case, an entire water system.
This type of assessment will drive decisions regarding the hardware platform that will be selected and the need for redundancy in equipment and programming, including processors, networks, and data storage devices.
Specific information that will flow into the system for control and data collection, including the system piping configuration (discharge and inlet as applicable), instrumentation including I/Os and their engineering units, operating parameters, and high/low ranges such as pressures, flows, analog values, etc., should be referenced. This is often accomplished through preparation of what is known as a Piping and Instrumentation Diagram (P&ID) (Figure 1).
SCADA System Components
All SCADA systems serve three main purposes and have four components designed for each purpose (Figure 2).
Data acquisition consists of field sensors and controls (i.e., instrumentation), including inputs from primary process sensors such as water level or pressure transmitters and outputs to controlled devices like valves and pumps.
These values are then transmitted to a local or remote-control panel, device, or station, generally referred to as a Programmable Logic Controller (PLC), Distributed Control System (DCS), or Remote Telemetry Unit (RTU).
Although similar in architecture, there are distinct differences between a PLC and DCS (Figure 3). PLCs are traditionally used for single batch or high-speed control with a relatively simple and low-cost design and function as the core of the control system. Their design is flexible and generic but completely customizable. Processing time for tasks is typically very fast; operators usually interact and control the
system using some sort of graphical display such as a SCADA monitor.
A DCS is used for continuous, complex controls and possesses an integrated control center much like SCADA, which generates the core of the control system versus the processors in a PLC system. Converting the data from field‐level signals and protocols to the SCADA communications protocol in use is typically done at the RTU before transmission to the SCADA servers or central master unit, often called the Master Telemetry Unit (MTU).
An RTU can consist of a basic interfacing device with input and output modules along with a transmission modem or a sophisticated device that includes the functionality of a PLC. These RTU types, often referred to as smart RTUs, are capable of independent programming to enable the assumption of critical control commands by implementing alternative programming in the event of master site or communications failure. Figure 4 illustrates an example of a smart RTU.
Data communication can consist of various data transmission methods, including FM radio, cellphone service, and dedicated hardwire, usually in the form of unshielded twisted pairs of copper wires, shared (multiplex) or dedicated telephone circuits, satellites, cloud communication, and fiber-optic cables.
It comprises the system’s backbone for the branch networks, RTUs, and other networked devices. A Human-Machine Interface (HMI), such as a touchscreen, keyboard, or mouse, is generally available at the master site to modify setpoints or collect/store data to or from the remote sites.
Data presentation relies on an HMI that includes the SCADA process servers; a data historian; and other ancillary devices such as firewalls, print servers, and thin clients throughout the facility. Thin clients, also called slaves, are stripped‐down computers that have no hard drive or installed software, running instead from a central server.
The RTU must be able to accept any standard control signal and convert it to the network protocol in use. This conversion occurs within the RTU, PLC, or whatever processor is being used and formatted for the data frame format. In this way, any valid signal can be monitored and subsequently transmitted to the SCADA master or server for processing.
The process is bidirectional (often called full duplex in communications language) in that control signals are also sent to the RTU for conversion and application in the field. These signals include pump start/stop commands, variable frequency drive speed commands, and control valve positioning, among other functions.
Any number of RTUs can be installed in the system, and each RTU can be configured to communicate with another, allowing an operator to access any other process through an operator interface terminal or external thin client.
Many modern SCADA systems, particularly radio systems, utilize polling interrogation methods between multiple sites. Polling is the process where the master computer or controlling device transmits and receives data from individual sites, using a sequenced rotation.
Multiple-site polling is an economic and efficient way to incorporate many remote sites into a single network without having to dedicate continuous transmission to the sites. Each time the site is interrogated, data from the site is sent and updated with new operational instructions to the site relayed.
A polling cycle is the time in which each element is monitored once. The optimal polling cycle will vary according to several factors, including the desired speed of response and the processor time and bandwidth of the polling.
In roll call polling, the polling device or process queries each site in a fixed sequence. Because it waits for a response from each site, a timing mechanism is necessary to prevent lockups caused by non-responding sites.
Roll call polling can be inefficient if the response time for the polling messages is too high; there are numerous sites to be polled in each polling cycle and only a few sites are active.
In hub polling, also referred to as token polling, each site polls the next site in some fixed sequence. This continues until the first site is reached, at which time the polling cycle starts all over again.
Polling has the disadvantage that if there are too many devices to check, the time required to poll them can exceed the time available to service the I/O device. Thus, prudent consideration must be applied when using a polling network on multiple sites. Figure 5 illustrates an example of a SCADA system using a series of polling radio signals to remote sites.
It’s obvious that data communication is an infrastructure that’s essential for the entire SCADA system to properly function and is how field data are transmitted from the RTU to the MTU or servers.
The RTU connects to the chosen communications medium and becomes a node that can transmit and receive data from the servers or from another RTU. Process networks are usually redundant, meaning there’s more than one communication path and each RTU has the capability of reverting to the other path in the event of a primary network failure, a technique known as self‐healing.
Network redundancy is the process of providing multiple paths for transmission of data, so that the data can keep flowing even in the event of a failure. Simply stated: more redundancy equals more reliability and assists with distributed site management.
The idea is that if one device fails, another can automatically take over. By adding a little bit of complexity, the probability that a failure will crash the network is greatly reduced. Figure 6 is an example of a redundant network.
An important new trend in SCADA systems is the use of virtual or cloud servers, which means there’s no physical enclosure where all SCADA functionality resides. Rather, the system is a software element that’s essentially a collection of files that functions on a specialized computer. There’s almost no limit to the number of virtual servers a computer can host.
Virtualization allows for enhanced security since the server’s access to the physical world can be controlled at a fundamental level on all sides of the physical computer. Older SCADA systems generally use a dedicated computer for each server, which is an expensive approach in terms of equipment and staff needed to operate and maintain the system. Multiple servers require multiple data connections, resulting in an overexpanded network infrastructure. With virtualization, one dedicated computer can host multiple virtual computers and servers, greatly reducing capital outlay and operating costs.
Another communication method is the use of wireless technology, which eliminates the expense of wiring while providing multiple signal paths between the RTU, the servers or MTU, and other RTUs. This is called mesh networking and is a powerful, reliable approach.
Using multiple radio frequency paths to communicate requires more computing power, but the security and reliability of a mesh network make it far superior to physical cables that have the unfortunate tendency to get damaged.
Using cellphones and internet technology is another way to transmit data and commands between sites. The use of selective polling allows commands to be transmitted to the sites and updates returned from the sites while avoiding excessive access charges. Figure 7 illustrates an example of a wireless SCADA system using internet access via cellphone transmission.
SCADA System Security
Security is another important aspect of designing and operating a control system, especially for computerized SCADA systems with remote access capability. Aside from virtualization, measures must be taken to protect the SCADA system from intrusion or compromise (i.e., hacking) by unauthorized individuals.
Only those with proper authorization should be able to access or modify a SCADA system’s functions. A hierarchical privilege system (Figure 8) is an essential and logical first step towards this objective. The principle of “least privilege” should be adhered to, allowing only those with a legitimate need and purpose to access the system’s most sensitive and potentially damaging elements.
Typically, at least four user levels are recommended in an advanced computerized SCADA system used for potable water supply control, with each level providing greater access and privileges than the previous lower level.
The first or lowest privilege is a way to allow routine access to only the information needed for users to perform their assigned tasks. This level is generally accessible by operators and permits routine access to all operational screens and system parameters and to make minor adjustments such as data value reset.
The second level permits a basic modification of process and operating setpoints, as necessary, for ensuring smooth continued operation of the system.
The third level usually permits a slightly higher level of access with access to configuration screens and basic control logic access, including modifying control algorithms and modifying and resetting adjustable control setpoints, alarm points, and data collection frequencies and ranges.
The fourth and most secure level is reserved for the control system vendor, architects, and programmers. This step requires the highest level of password security to enter the system and permits advanced changes to the control and programming logic, fixed and advanced program instructions, and security codes.
When remote access to the system is enabled, the third and fourth levels should not be accessible through remote means and reserved for onsite access only. This prevents possible hacker access that could modify pump operation, water level and pressure values, and chemical feed rates. Provisions should be added to enable locking out or disabling unauthorized remote users or unsuccessful password entries.
SCADA System Control
One of the most important components of any SCADA system are the instrumentation and control devices that directly interface with the local processes. These devices are how the process is monitored and controlled. Without these devices, the communication and presentation components would have little or no purpose. This element of a SCADA system deserves a separate discussion and will be outlined in the April installment of The Water Works.
SCADA is a powerful technology and tool, but water system managers and operators along with SCADA system designers need to follow best practices to take full advantage of its strengths and capabilities along with recognizing its limitations and potential drawbacks.
The success of any control system lies in the ability for the operator to access and transfer valuable data needed to safely and efficiently operate the water system. Ultimately, this data must be presented to operators for observation and action in a timely manner.
Typically, the most visible elements of any SCADA system are the graphic display screens depicting the process being monitored and controlled. These screens show how an operator maintains control of a water system. When the RTU transmits field data back to the SCADA servers, specialized software uses this data to activate a display screen element or status and data. Equipment status, process variable values, and alarm rates can be displayed on these screens.
The software can manipulate screen elements to provide an operator with a clear and concise picture of what’s happening in real time. Each data point is listed in a database that allows the software to link the field data to points on the screens.
Key Performance Indicators or KPI screens are often used to show operators the health of their systems at a glance. This has become a popular method of process monitoring and is rapidly replacing the use of multiple process screens and annunciators.
Without KPIs, a SCADA system might use multiple graphic screens, usually one for each process area. But by using a KPI screen, an operator can quickly determine where a problem exists or simply zero in on a particular process, scrolling down to the particular data point required. Each screen element has a pop‐up context menu that allows an operator to modify equipment operation or adjust setpoints.
All data presented on the screen are recorded and archived for future reference by the water system or regulatory authority. An example of a complete SCADA system with both radio and landline remote access to the RTUs is shown in Figure 9.
______________________________________________
This concludes this installment of The Water Works. In April, we will continue this discussion on advanced control systems with an overview on instrumentation and control system devices.
Until then, keep them pumping!
Ed Butts, PE, CPI, is the chief engineer at 4B Engineering & Consulting, Salem, Oregon. He has more than 40 years of experience in the water well business, specializing in engineering and business management. He can be reached at epbpe@juno.com.