Sep-2020

Planning to counter economic turbulence

Synchronising engineering and planning models should be a key aim to support the resilience and agility needed for economic recovery

RON BECK
AspenTech

Article Summary

Global economic disruption has upended the refining industry. Not only are the dynamics of the situation difficult, but they seemingly change on a weekly or even daily basis.

In this territory of future price and demand uncertainty, refinery planners and schedulers are in the hot seat. Refinery executives and managers are forced to examine new scenarios daily, and each scenario needs to be evaluated for safety, logistics, and economics. The organisation’s planners are at the nexus of this.
They will be the key to a refinery’s economic resiliency and agility in the future. They require advanced technology to ensure the accuracy of their digital models in non-standard operating regimes, and ability to examine many scenarios quickly to optimise across a wide envelope of options and opportunities.

The business challenge
Refineries are typically set up to produce predominantly a mix of gasoline, diesel, naphtha (as gasoline and petrochemicals feedstock), and jet fuel. In today’s market, with gasoline and jet fuel in slowly recovering demand and over-supply, and diesel the preferred transportation fuel output, planners and engineers are being asked by executives to rapidly develop refinery plans that maximise diesel. In tomorrow’s market, with the energy mix in transition, most rapidly in Europe, planners can be strategic in advising executives as to the best capex and process strategies to achieve agility and resilience in the face of uncertain future market evolution. Both questions are being asked now.

Today, a wider range of crudes are available at low prices, which can be enticing, but future demand and pricing are uncertain, and refineries can potentially lose a lot of money if wrong choices are made. With diesel prices holding better than other product prices, maximising diesel is, in the short term, a key objective. Jet fuel demand is likely to be the slowest to come back. This focuses attention on several key refining units, especially the CDU and VDU, where the fractions that yield diesel can be maximised. Further, with maintenance crews having been largely taken out of the asset for health reasons, the integrity and safety of the process units must be reassessed when flows and parameters are changed significantly. This calls for engineering advice relative to scenarios that may look advantageous to planners.

It is now clear that assets may need to be operated in a new normal, with lower staff density on site. This requires looking at the interaction between production, catalyst lifetime, and asset health and integrity.

There are many cascading, related refining operational questions that planners and engineers are asked to answer.  What is our lowest throughput safe operating level? Should we order catalysts early due to supply chain interruption? Can we make minor process reconfigurations to better utilise our intermediate products? What other moves can we make to take advantage of crudes that may be available at lower prices? Which operating plans give us better flexibility in face of extreme volatility of prices, supply, and demand?

Collaboration to respond to the challenge
Close collaboration between planners and process engineers is required to answer these questions. A digital solution is the best enabler of rapid response to this challenge. The modelling systems (for instance, the Aspen PIMS submodels used in about two-thirds of all refineries) used in planning a typical refinery were not originally implemented or tuned for the kinds of eventualities being encountered dynamically today. Before the scenarios can be run with accuracy, the models of these key economic units must be changed to reflect accurately the proposed operating regimes. The engineering digital twin models of key units are a crucial competitive advantage to obtain a fast, actionable answer. And an automated workflow to enable the engineering model prediction to inform the planning submodel, without significant manual time, provides a critical advantage.

Figure 1 shows the collaborative work process required to respond rapidly to the need to scenario plan in this environment of volatility, uncertainty, complexity, and ambiguity.

The following are typical steps needed to answer executives’ questions about scenario planning for increased diesel, low jet demand, rapidly changing prices, and flexibility for the future.

Address dramatically different product mix scenarios (maximising diesel and minimising jet and gasoline cuts; be prepared for increased naphtha demand as olefin feedstock).

The CDU and VDU submodels most importantly, but also hydrocracker, hydrotreater, and/or FCC submodels in the planning model were most probably not developed with the requirement in mind to look at these ‘extreme’ situational plans with any accuracy. When these linear submodels are pushed to outside the expected limits, the accuracy of the prediction will go way down. The scenarios are outside the existing model’s range of accurate predictions. These model elements need to be quickly and efficiently rebuilt based on a new range of expected operating limits.

First, the simulation (Hysys Refining or similar) models need to be quickly updated, to match current operating data, and the process alternatives for maximising diesel cuts and minimising jet and gasoline cuts need to be created and modelled. Working remotely, as most are today, the engineer sets himself up for cloud access to the models, then obtains current data sets to recalibrate and update the model. The calibration can be accelerated using today’s technology, with recent innovations in performing calibration assisted by data science, step-by-step widgets, and in-software advice. This needs to be done with constant collaboration with the planner while working remotely. By running a sensitivity analysis of the planner’s desired options, Hysys will inform the safe operating limits for the CDU and VDU and physical properties that need to be used for these new operating regimes.

Next, those CDU and VDU model results must inform an updating of the planning submodels. This can be done by communicating to planning the new safe operating limits, correct assay information, and new base delta vectors for the PIMS-AO (or similar) submodel. AspenTech has available a convenient and efficient workflow for updating the various PIMS submodels and is progressively automating that.