Sie sind auf Seite 1von 5

WHICH CONTROL OPTION DO I GO WITH? DCS, MINI-DCS OR PLC?

Kevin Stively Senior Process & Controls Engineer Parker, Messana & Associates 4507 Pacific Hwy. E., Ste A Tacoma, WA 98424 ABSTRACT In todays world of automation, you now have many cost effective control solutions that allow for advanced computer control. With these advanced control solutions come the gigabytes of process data that can be utilized for quality control, process optimization and maximizing equipment capacity. Each type of system has different costs associated with their implementation. Many of these costs are associated with proprietary, semi-open control solutions. With advances in PLCs, personal computers and operating systems, end users now have the ability to install cost competitive control solutions while maintaining the quality of control and data availability. The base process to be used in the control solution comparison is a vacuum washer line with knotting and screening. The three systems solutions for the comparison will be a traditional DCS, mini-DCS and PLC/PC (SCADA control).

I. OVERVIEW More control options exist today than did in years past. Ever since technology allowed industry to pneumatically and electronically manipulate a process, engineers have been on a quest for optimizing and balancing the cost to quality equation. Improving product quality can be achieved in many ways. One of which is through better control. Over time, consistent quality of a product can be best achieved by using a computer in one form or another. It could be a local single-loop controller, multiple single-loop controllers, fieldbus control or a central controller such as a DCS. Management is faced with cost decisions every day. Wouldnt those decisions be better made from more information? DCS, Mini-DCS and PLC/SCADA (Supervisory Control and Data Acquisition) control systems offer all of the relevant process information in a single user format (using excel or other database software). If you already have a DCS, but it is really old and not Y2K compliant, do you want to spend thousands or perhaps millions of dollars to upgrade it or replace it with some other technology? On the fast road to better control technology, one can not make decisions based on past performance, but rather what current technology has to offer. It might even be more cost effective to update or replace less expensive, but equally capable control systems every five years than to invest in a mega-control system that may have a 10 year ROI. If you are looking to upgrade an existing control system or install one for the first time, you may be asking yourself Which control system/technology should I go with? This paper might help you in your quest. II. PROCESS BASIS To create a realistic control solution comparison, a vacuum washer line with knotting and screening will be the process example. The washer line is a four stage cascading wash sequence with individual liquor tanks and cascading shower liquor for optimal pulp washing. Primary and secondary pressure knotters along with three stages of pressure screening prepare the pulp for the next stage in the process, which is bleaching.

1999 TAPPI. Reprinted with permission, from Conference Record of the TAPPI 99: Preparing for the Next Millennium, PCE&I Joint Conference; Atlanta, March 1-4, 1999.

This process unit contains the following I/O: 110 analog inputs 68 analog outputs 40 discrete inputs (non-motor) 15 discrete outputs (non-motor) In addition to the aforementioned I/O, are 50 motors with three discrete inputs (ready, running, and overload) and one output (run) to be controlled by the control solution. Twenty-four PID Loops are required to control the process in this example. III. CONTROL PLATFORMS Three control platforms will be discussed for control of the example process. They are a traditional DCS, Mini-DCS and PLC/SCADA/HMI. Each platform has advantages and disadvantages in regards to design, costs, benefits, and limitations. Each of the control systems will contain three 20-inch touch screen operator consoles, one engineering station, 15 kVA UPS, color printer, and dot matrix printer. IV. DESIGN CONSIDERATIONS Each of the three systems has different design considerations. The traditional DCS will use a PLC for motor control. Although this is not necessary, many mills choose to have motor control in a PLC. The DCS will also have a redundant processor and communications module to the PLC (again, this is not required, but customary in large systems). Two processors are required for the Mini-DCS and PLC systems to provide enough control capacity. The Mini-DCS is limited to 250 I/O points per processor. The PLC could handle all of the I/O and control loops, but the scan time would be slower than desired so two PLC processors will be utilized. This will also allow enough capacity in the PLC for advanced calculations if required. Typical scan times with PID control are in the 20 to 60 ms range. Each of the three systems will be sized with 50% open spare I/O rack capacity and 15% installed spare I/O. While 50% spare rack space may seem excessive, it eventually will be used with future control projects or process additions. In addition to the hardware aspects, one needs to consider the network architecture and operating environment, safety/software interlocks, and user requirements such as graphics, operation, reporting, management and regulatory needs. For this example process, each of the systems will have twelve control graphic pages, five help screens, ten trend pages, and an alarm summary. A daily summary of key operating data will be generated automatically (or on command) for management review once per day before the morning production meeting. Another consideration is the amount of time and resources it takes to configure each of the systems. More engineering and configuration time may be required to install and configure a PLC/SCADA/HMI system than a Mini-DCS and DCS. Most of the time difference is in the configuration of the HMI and advanced control strategies. Some DCS systems may be more complicated than others to configure, so careful review of the time required for each type of system needs to be included in the cost of a control project. The cost of the control room is not included in the design basis and is considered equal in all three of the control needs. For sake of simplicity, it is assumed that the I/O rack room exists and is large enough for the largest I/O footprint, which is the DCS. Considerable cost savings could be factored into the overall system if space is limited and a new rack room is not required due to the smaller footprint systems like the Mini-DCS and PLC/PC systems. The cost of training is also not included in the cost comparison. Training is a variable that each site much weigh against the level of experience their staff has.

V. COST COMPARISON The DCS is chosen for comparing the systems to each other, since it usually is the more expensive of the options. The comparative cost of the Mini-DCS is 48% less than the DCS. A PLC/PC system costs 44% less than the DCS used in this example. Figure 1 graphically shows the cost difference of the systems by physical (real) I/O. The cost line for the Mini-DCS and the PLC/SCADA are virtually the same. So only the PLC/SCADA line is displayed on the chart. The costs of the DCS approach the cost of the PLC/SCADA when the point count is greater than 1200 physical I/O. The example process requires 433 physical I/O points. Each system must contain 15% installed spare for the cost comparison, which brings the installed I/O count to 496 points. The cost per point (with 496 I/O points installed) for a DCS is $1900, Mini-DCS is $1047 and $1085 for a PLC/SCADA/HMI system. VI. BENEFITS COMPARISON The DCS is the clear choice when it comes to large-scale process control needs. It has many advanced control algorithms that PLC/SCADA systems do not. A DCS has more capability for long term data manipulation over a PLC and Mini-DCS. However, this is changing. It took years for todays DCS to reach the level of utility it has. PLC/SCADA combination and Mini-DCS systems have only been around for the last few years. In time, they too will have the same level of capability. The Mini-DCS is the scaled down/simplified version of the large scale DCS. Manufacturers of traditional DCS systems have created a scaled down DCS, which in many cases resembles a PLC with one big difference. The MiniDCS has all of the history in advanced control technology and engineering tools from the learnings of their larger DCS cousins built into the system. The Mini-DCS is a less capable system in comparison to a DCS, but many industrial processes can be controlled very nicely with the Mini-DCS. The only draw back is the firmware limitation on amount of I/O per processor. The Mini-DCS has a small footprint like the PLC.
Cost Comparison of Control Systems Overview
$40,000

$35,000

$30,000

$25,000 Cost Per Point

$20,000

$15,000

DCS

$10,000

$5,000 PLC/SCADA $16 256 496 Physical I/O Points 736 976

Figure 1. Cost Comparison of Control Systems Overview

PLCs offer the most flexibility and scalability of the three systems. The PLC/SCADA/HMI systems are more open in that they use standard personal computers for the operator interfaces and TCP/IP (Transaction Control Protocol/Internet Protocol) over standard Ethernet for networking. Other users on the network can query data from the SCADA node or directly from the PLC using DDE (Dynamic Data Exchange) capable software. Another advantage of PLC/SCADA control systems is that the SCADA node can usually communicate directly to and between different DCS systems. The PLC typically has more types of I/O cards available than the Mini-DCS. VII. LIMITATIONS The only limitation a DCS has is it ability to interface with other manufacturers DCS systems and exchange data through standard networking means (i.e. Ethernet, which is what PCs use). While this problem is slowly being addressed, it has taken over 15 years to solve. Mini-DCS systems offer many of the benefits mainstream DCS systems offer. They have great configuration tools built into the processor. This may be considered a limitation to the extent that you have to train people to be able to configure the system. However, most of these newer configuration software are very user friendly and utilize standard Microsoft Windows features and techniques. Another major limitation of many Mini-DCS systems, is the availability of different types of I/O. They often offer standard 4-20 mA I/O and 120 VAC discrete I/O but no other specialized I/O. This problem is viewed as a short term one since Mini-DCS systems are relatively new to the market. As demand for the specialized I/O is made, the manufacturers will make them available. VIII. CONFIGURATION All three of the systems require configuration. Each, however, require different levels of knowledge to be able to configure a process with advanced control needs. Most DCS systems require more initial training than the other two. The DCS and the Mini-DCS both have great engineering tools built into the system for configuration. This is because the mainstream DCS suppliers are the manufacturers of the Mini-DCS. In addition to control configuration, operator interfaces require configuration. The software is usually limited to the manufacturers configuration software. PLCs have many third party configuration software packages available. These software require minimal training, but are PLC specific. Additional software is required to configure the SCADA and HMI aspects of the control system. Again, many software products are available to create great operator control interface screens. IX. TRAINING REQUIREMENTS Training can be a significant part of a control system. DCS systems require more up front training because of the nature of their design. They have multiple components that all interface together forming the control system. If you skimp on the training, it will only cost you more in the end in lost time and staff frustrations. Mini-DCS system will require training from the view that they are new to the market with little exposure to your staff (unless you already have one installed on site). PLCs have been around for many years and probably do not require much training to support. The software used to configure the PLC and the SCADA/HMI may require training if the staff has never used this type of software in past projects. Operator training for graphics will familiarize them on how to control the process though the new interface. Operators who have only controlled the process through local, panel and pneumatic controllers may require additional training on how the new control system works. This will help with comfort levels. It is also advantageous to include them in the graphics design instead of presenting them with a completed operator station. This allows for operator input for their needs instead of what someone else thinks their needs are.

X. THE FUTURE OPC, fieldbus, fuzzy logic, and other advanced control software will elevate PLC/SCADA/HMI control systems to the same capability as a DCS has today but at a much more cost effective level. New hardware developments such as USB ports for PC will offer new ways to interface computers directly to the process and potentially replace PLC/PC combinations for PC running real-time control software (which is a whole other paper). XI. CONCLUSIONS Large processes with more than 2000 physical I/O point and a few hundred PID loops might be best served by the traditional DCS. However, if the process is not complicated a Mini-DCS or PLC control system will work well. The cost difference between the Mini-DCS and the PLC are minimal. When determining which system is the best for the process, one must consider what kind of I/O is required and how advanced are the process control needs. A Mini-DCS might provide better tools for advanced control needs, but has limited types of I/O. The clear choice for a process with a moderate I/O count, of which many are discrete, and a moderate number of PID control loops is a PLC/SCADA/HMI control system. It offers the most flexibility and is modular in its implementation.