Stand-alone solutions needed
In what circumstances can users do without redundant pumps at large chemical plants?

The need to reduce investment costs while maintaining high availability inevitably raises the following question in the chemical industry: where and under what conditions can users do without redundant pumps? Maintaining conformance to stringent product quality standards is not the only issue. Users also need suitable monitoring equipment and early warning systems.

The changing world of chemistry: certain products can only be produced cost-effectively at world-scale plants. We are also seeing a tendency to build single-stream plants rather than plants that operate parallel production lines. This is done to reduce investment costs. Many production lines are embedded into a series of upstream or downstream systems to exploit synergy effects. This results in larger flow volumes and the need for maximum availability. Because pumps are becoming larger and more expensive, manufacturers and users need to work together to find answers to the following questions:
-Why are redundant pumps used despite the fact that single compressors are used on the same system?
-What conditions are necessary to eliminate the need for redundant pumps?
-What options are available to increase pump reliability?
-What warranty terms and availability guarantees are producers prepared to offer?
Rethinking conventional pump strategies
The following strategies can be used: (A) reserve pump strategy: “100% pump” with 100% redundancy, “50/50% pumps” with 50% redundancy; (B) single pump strategy: “100% pump“, no redundancy; (C) multi-pump strategy: “60/60% pumps“, no redundancy. Past experience with the production process, the pump type and the pump manufacturer is the crucial factor which influences the selection of the redundancy strategy. The following solutions may be appropriate:

-safety-related pumping application (e.g. reactor cooling circuit): redundancy is a prudent solution;
-new “unknown” process (e.g. pilot or research center system): redundancy is a prudent solution;
-established” process (a number of systems have already been built): there may not be a need for redundancy;
-a certain amount of down time is acceptable (e.g. one shift): there may not be a need for redundancy;
-no references for the pump type or manufacturer available: redundancy is a prudent solution;
-associated auxiliary systems: redundancy is a prudent solution.
Regardless of the pump strategy used, a detailed discussion between the user and the manufacturer should take place in cases where there is a need for high availability. The two sides should review the process and system requirements and the resulting stress which components and assemblies will have to withstand.
The discussion should include the shaft seal, the type of bearings, the materials used and auxiliary systems. If redundancy is used, the only thing needed may be a list of spare parts. For single pump solutions, the parties may want to develop a defined quality assurance and commissioning strategy. If the plan is to use a “100% pump” with no redundancy, the anticipated process flows should be specifically tested during commissioning to assess the process relevance of the pump (the design basis). The effect on the shaft seal and bearings can be determined depending on the operating points, state values and material properties of the medium. Axial and radial forces vary as the pump’s operating points change, and this alters component stress which is caused by forces, moments and vibration. Output and efficiency change with the flow, and this affects the temperature distribution inside the pump. There are also changes in pressure, temperature, viscosity, homogeneity (gas, cavitation and solids), susceptibility to corrosion and abrasion, explosivity, toxicity and changes in fluid state values and material properties.
Verification of process capability during pump commissioning gives the user the opportunity to upgrade the pump, and the pump manufacturer can make any improvements that might be necessary.
The life cycle costs should also be assessed
When the decision is made to use a single-pump strategy rather than a reserve or multi-pump strategy, particular attention should be paid to the cost of starting up and shutting down the system during ongoing operation and the cost of lost production that could result from a fault. This must be compared to the cost of one or two reserve pumps. The life cycle costs should also be assessed, but the relevant data will have to be generated by the user and the manufacturer. It is not meaningful to make a general statement about which pump strategy is best in any particular application even when data from previous systems is available. Various factors such as system size, integration at the site, external conditions, qualification of operators and maintenance personnel and on-site maintenance and repair capability vary from location to location, and these factors can lead to different results and strategies.
A quality assurance program which includes production of the pump and purchased parts is essential to meet stringent quality and availability requirements. Especially in applications without redundancy where large, expensive pumps with long lead times are used, quality assurance plays a vital role. Quality can be divided into the quality of the pump itself, the quality of the process data and the quality of machine diagnostic/early warning systems. Process and design data must be reproducible and be safeguarded in the system to ensure fault-free pump operation.
Building on established quality systems such as ISO 9000, the user and the manufacturer must carry out specific quality planning for the particular application to ensure compliance with the enhanced pump reliability requirements. This is especially important for functional and safety-related parts. Careful attention must be paid to the pump housing production process at the foundry or forge. It is important to ensure that state of the art methods are used and that all relevant material know-how is reflected in the production process. Besides a pattern which is optimally adapted to the material and the workpiece geometry, a simulation of microstructure solidification is an important quality assurance tool. It is essential that heat and surface treatment and production welding are carried out properly. Final, non-destructive testing of the complete workpiece must also be performed. To ensure quality, the pump manufacturer will have to increase his quality surveillance activities at his suppliers’ sites to ensure that specifications are being met, and this requires trained personnel.
For the reasons mentioned above, it is essential that only qualified suppliers and manufacturers with the appropriate experience, capabilities and quality management and assurance systems are
selected. This presupposes detailed knowledge about manufacturer/supplier work methods and flows, and it includes auditing and validation. There is no other
way to ensure that pumps are made under similar conditions to compressors, which helps eliminate the need for redundancy.
Monitoring and early failuredetection
If users choose a single-pump strategy, they need to monitor the pumps and assess operational status. Machine diagnostics and early failure detection systems are suitable tools which enable users to perform status-based maintenance to increase availability and limit financial losses caused by production stoppages. They also facilitate intelligent stock level strategies. Monitoring based on machine diagnostics and early failure detection
also reduces the risk of production losses on systems that use a multi-pump strategy without redundancy. If pump load and stress factors are known, the pump is properly designed and suitable materials are used, it is reasonable to expect that pumps which have a nominal design life of 15 years will run fault free for at least five years in the process industry. Pump failure during this five year period can only be the result of early failure, random failure or failure caused by operating conditions outside the specified limits.
Failure can also occur when pumps begin to wear out (end of service life). Expert systems are available for large pumps which use vibration (acceleration, displacement and velocity), electrical parameters (current, voltage, etc.) and process parameters (temperature, pressure and flow) to identify operation outside the allowable limits or detect end of life on high wear parts. These systems provide timely, reliable, differentiated information. It is also worth mentioning adaptive control. Features such as speed control can be used to keep unplanned stress within allowable limits.
Summary: If pumps are installed at
large chemical plants in non-redundant configurations, machine diagnostics and early warning systems must be put in
place to monitor and assess operational status. In combination with professional risk management and excellent quality management, this approach can guarantee high system availability.