Refinery Production Simulation: Production vs. Realization

       Refinery production simulation can evaluate a refinery from two main perspectives: production volume and product realization.
       Production-based refinery assessment focuses on the production volume generated by the refinery’s processing units. For many refinery operators, this is the primary measure of production performance.
       Realization-based assessment considers the complete path from production to shipment. In addition to production volumes, it accounts for product inventories, tank farm capacities, pumping and loading rates, loading racks, shipment operations, and transportation constraints. A realization-based assessment evaluates whether the products generated by a refinery can actually be stored, moved, and shipped through the downstream system.
       A refinery simulation model may need to support both assessments. The challenge is to do this without maintaining two separate model architectures. With Petroleum Refinery Library, this can be achieved by building the complete production-and-realization architecture once and configuring its downstream elements so that they do not constrain production when a production-only assessment is required. The following sections describe this practical approach and show how to switch between the two assessment modes without rebuilding the model.

       Production-based assessment focuses on the output of the refinery’s processing units under specified operating conditions. It is particularly useful for evaluating:
        - refinery production planning and plan execution;
        - refinery throughput and production capacity;
        - processing unit utilization;
        - product yields and production volumes;
        - the impact of feedstock constraints and operating conditions on production output.
       In an integrated refinery simulation model, production is connected to downstream systems such as tank farms, loading racks, and transportation facilities. Their capacities and operating constraints can therefore become downstream bottlenecks that artificially limit refinery production. The practical challenge is to retain these elements in the model while preventing them from constraining the production assessment. This allows the same model architecture to be used for both production and realization analysis.

       A product realization assessment evaluates whether products generated by the refinery can actually be stored, transferred, loaded, and shipped through the downstream system. It therefore extends the production assessment to include storage, shipment, and transportation. In a simulation model, this requires representing the main constraints along the product flow:
        - initial product inventories;
        - tank farm capacity;
        - pumping and transfer rates;
        - loading rack capacity;
        - shipment operations;
        - rail and road transportation availability;
        - transportation and loading schedules.
       These constraints interact dynamically. A full tank can prevent further production from being transferred, limited loading capacity can delay shipments, and insufficient transportation availability can cause products to accumulate in storage.
As a result, realized volume can be lower than production capacity, even when the processing units themselves have sufficient capacity. This makes realization-based simulation valuable for assessing the refinery as an integrated production and logistics system rather than evaluating processing units in isolation.

       Supporting both assessment methods may suggest building two separate architectures: one for production and another for production and realization. However, this approach creates unnecessary complexity. The same tank farms, loading racks, shipment nodes, and material-flow algorithms must behave differently depending on the assessment objective. In production mode, downstream constraints should not limit output; in realization mode, those constraints must be fully represented. Simply switching individual objects to Direct Flow is not sufficient. Other model components may depend on their normal storage or shipment behavior, so changing one object can affect the logic of connected algorithms. Maintaining two architectures also means duplicating development, testing, and future changes. A better approach is to build a single refinery simulation model covering both production and realization and switch its operating behavior according to the assessment mode.

       The most practical approach is to build a complete refinery production simulation model covering the full production-and-realization process from the beginning. The physical structure of the model remains unchanged. Instead, several parameters and operating modes are adjusted:
       1. Loading racks are switched to Direct Flow.
       2. Initial tank farm inventories are excluded from the calculation.
       3. Tank capacities are increased where necessary so that storage does not limit production.
       4. Production statistics are collected upstream of the tank farms.
       5. Shipment and other downstream constraints that could restrict the flow are removed.
       Together, these changes transform the downstream part of the model into a non-constraining flow path. Products can continue through the existing architecture, but storage and realization constraints no longer determine the measured production volume. This approach preserves a single model architecture while allowing the same simulation to answer two different questions: how much the refinery can produce and how much of that production can actually be realized.

       Loading racks typically have little influence on upstream production. Their primary role is to transfer finished products from the refinery’s storage system to the transportation system. For a production-only assessment, this realization step is unnecessary. In Petroleum Refinery Library, loading racks can therefore be switched to Direct Flow using a Boolean parameter that defines the selected assessment mode.
       Conceptually:

Production + Realization → Normal loading rack operation Production only → Direct Flow

       The loading rack remains part of the model, but it no longer represents a shipment operation that can restrict the upstream material flow. This is the simplest of the required changes. Tank farms require more careful treatment because their inventory and capacity can directly affect the production process.

       For a production-only assessment, initial tank farm inventory should not contribute to the measured production volume. Otherwise, the material balance may combine newly produced volumes with products already present in storage. In Petroleum Refinery Library, this can be controlled through the corresponding setting in Constants, which allows initial tank farm inventory to be excluded when the model is initialized. This isolates the production generated during the simulation period. However, tank capacity can still restrict the material flow, so the next step is to remove storage capacity as a potential production bottleneck.

       Even without initial inventory, a tank farm can still limit production when its available capacity is insufficient to receive additional product. For a production-only assessment, this constraint can be neutralized by assigning a sufficiently large capacity to the affected tanks. In practice, the affected tanks can be assigned a capacity large enough to prevent them from becoming full during the simulation. This does not require changing every tank farm. Only storage locations whose actual capacity could restrict the production flow need to be enlarged. The result is an effectively non-constraining downstream flow path. The tank farm remains part of the model architecture, but its storage capacity no longer determines production output.

       Changing tank farm capacity is not enough if the production KPI is still collected downstream of storage and shipment operations. For a production-based assessment, the production KPI should be measured upstream of the tank farm, at the point where the product enters the downstream storage and realization system.
For example:

Processing Unit → Production Statistics → Tank Farm → Loading Rack → Shipment

rather than:

Processing Unit → Tank Farm → Loading Rack → Shipment → Realization Statistics

       The final step is to eliminate any remaining downstream constraints that can restrict refinery production flow.
Depending on the model, these may include:
        - maximum shipment volumes;
        - pumping or transfer limits;
        - loading capacity constraints;
        - other capacity constraints between the processing units and the production statistics point.
For a production-only assessment, these restrictions should be disabled or increased sufficiently so that they cannot create backpressure on the upstream process. After this configuration, the entire downstream section effectively acts as a non-constraining flow path. Products can continue through the existing tank farms and loading infrastructure, while their capacities and shipment limitations no longer affect the measured production volume. The physical architecture remains unchanged; only the constraints that are relevant to realization are neutralized.

       After these changes, the same refinery simulation model can support both production-based assessment and product realization analysis. In realization mode, the model uses the actual downstream system:

Production → Tank Farm → Loading Rack → Transportation → Realization

       Storage capacities, inventories, loading operations, and transportation constraints can therefore influence the realized volume. In production mode, the same structure is configured as a non-constraining downstream path:

Production → Direct Flow → Non-Constraining Storage → Statistics

       Initial inventories are excluded, relevant storage capacities are increased, downstream constraints are removed, and production statistics are collected upstream of the realization layer. The downstream objects are neither deleted nor replaced with a second model. They remain part of the architecture but no longer constrain the production KPI.

       When adapting a refinery simulation model for production-based assessment, several seemingly simple changes can lead to incorrect results.
       Building a second production-only architecture. This measures transferred or realized volume rather than production output.
       Keeping initial tank farm inventories in a production assessment. Existing stocks can become part of the calculated material flow and distort the production result.
       Increasing every tank farm capacity unnecessarily. Only storage locations that can constrain the production flow need to be enlarged. Other tanks can retain their actual capacities.
       Leaving shipment constraints active. Maximum shipment volumes, loading limits, or transfer restrictions can create artificial bottlenecks even when the assessment is intended to measure production.
       Building a second production-only architecture. Duplicating the model increases development and maintenance effort and creates a risk that the two versions will gradually diverge. The more robust approach is to keep one physical architecture and control which constraints are active for each assessment mode.

       Petroleum Refinery Library (PRL) provides the building blocks for implementing both production assessment and product realization within the same refinery simulation model.The key is to treat production and realization as two operating configurations of a single material-flow architecture. The refinery simulation model can use a Boolean parameter to select the production or realization assessment mode and apply the corresponding configuration during initialization.

Conceptually:
Production + Realization → Normal tank farm operation → Normal loading rack operation → Actual shipment constraints → Realization statistics Production Only → Exclude initial inventory → Increase relevant tank capacities → Switch loading racks to Direct Flow → Remove downstream flow constraints → Collect production statistics upstream

       This configuration-based approach avoids duplicating the refinery topology or production logic. Changes to processing units, material flows, or production algorithms can therefore be made once and used in both assessment modes. The result is a more maintainable simulation architecture that can evaluate both refinery production capability and product realization performance within the same model.

       One refinery simulation architecture can support both production assessment and product realization analysis. The production-only mode is obtained not by removing the realization layer, but by neutralizing the downstream constraints that could affect the production KPI. This preserves the model structure while allowing the same Petroleum Refinery Library simulation model to evaluate refinery production capacity and product realization performance.

1. What is the difference between production-based and realization-based refinery assessment?
Production-based assessment measures the volume generated by the refinery’s processing units. Realization-based assessment additionally considers storage, loading, shipment, and transportation constraints that determine how much of that production can actually be delivered.

2. Why can realized volume be lower than refinery production volume?
Realized volume can be lower when downstream constraints limit product movement. Typical bottlenecks include insufficient tank capacity, loading rack capacity, pumping rates, shipment schedules, or limited rail and road transportation availability.

3. Does a refinery simulation model need separate architectures for production and realization?
No. A single production-and-realization architecture can support both assessment modes. For production-only analysis, downstream constraints can be configured so that they do not restrict the measured production volume.

4. How can a tank farm be prevented from limiting production in a simulation?
Initial inventory can be excluded, and the capacity of tanks that could become production bottlenecks can be increased sufficiently. This makes the tank farm an effectively non-constraining downstream buffer for the production assessment.

5. What is Direct Flow in refinery simulation?
Direct Flow allows material to pass through an otherwise intermediate storage or handling element without normal storage or shipment constraints. In a production-only assessment, loading racks can use Direct Flow so that realization operations do not restrict upstream production.

6. Where should production statistics be collected?
For production-based assessment, statistics should be collected upstream of tank farms and other realization constraints. This ensures that the KPI represents production output rather than the volume that was ultimately transferred or shipped.

7. Should all tank farm capacities be increased for production assessment?
No. Only tanks whose actual capacity could constrain the production flow need to be enlarged. Other tank farms can retain their actual operating capacities and configuration.

8. Can the same Petroleum Refinery Library model be used for both production and realization analysis?
Yes. By configuring Direct Flow, initial inventories, tank capacities, statistics collection points, and downstream constraints according to the selected assessment mode, one Petroleum Refinery Library model can support both production-based and realization-based analysis.