# Refinery Dynamic Deadband Adjustment

*/Opportunities/Refinery_Dynamic_Deadband_Adjustment*

## Opportunity Overview

**Wedge**: The beachhead targets fluid catalytic cracking and coker units within mid-sized independent refineries. These specific units process highly abrasive media, causing severe valve erosion and making the financial pain of static deadband wear immediate and highly measurable. Expansion proceeds laterally to standard distillation columns before capturing full-plant PID loop coverage.
**Timing**: Sub-second process historian data now reliably streams to cloud environments, unlocking the ability to train ML models on massive time-series datasets. Simultaneously, reinforcement learning architectures are now stable enough to manage the non-linear dynamics of fluid systems without risking plant instability.
**Why This I C P**: Continuous process chemical plants run constantly with high financial penalties for unplanned downtime caused by single-valve failures. They possess the highest density of actuated control valves and already maintain the underlying sensor infrastructure required to feed dynamic models.
**Size Of Prize**: Approximately 2,000 global large-scale refining and petrochemical facilities experience roughly $300k per year in avoidable valve maintenance costs and deadband-related yield loss, creating a $600M annual market.
**Gap Narrative**: Refineries rely on static deadbands for thousands of PID control loops, forcing a permanent trade-off between tight process control and accelerated valve wear. Operators lack the analytical capacity to continuously tune these limits as feedstocks and ambient conditions shift. This rigid approach causes premature physical valve failure and suboptimal product yield.
**Defensibility**: Defensibility relies on high switching costs through deep integration into the facility distributed control system. As the software tunes more loops, it builds a localized dataset mapping specific valve actuation signatures to physical wear rates. Ripping out the system forces the refinery to abandon these highly specific, unit-level tuning models and return to blind, static baselines.
**Why This Thesis**: Software is the exact mechanism required to execute continuous, closed-loop mathematical optimization across thousands of variable tags. It integrates directly into existing Advanced Process Control systems to adjust parameters in real time without requiring physical hardware modifications.

## Opportunity Linked Thesis

**Thesis**: [Software](/Theses/Software)

## Opportunity Linked I C P

**Icp**: [Oil Refinery](/CompanyTypes/Oil_Refinery)

## Opportunity Market Sizing

_Illustrative — target and order-of-magnitude estimate figures, not an achieved track record (this Thing is concept-stage)._

**S A M**: ~$120M-160M (North American and European high-complexity refineries)
**S O M**: ~$5M-15M
**T A M**: ~1,000 global oil refineries × ~$300k-400k/yr per facility for control loop optimization software ≈ ~$300M-400M
**Growth Rate**: ~8-12%/yr, driven by the retirement of senior process control engineers and margin pressures on aging infrastructure
**Paid Comparable Spend**: ~$200k-500k/yr per facility spent on outsourced PID tuning consultants, legacy static APC software, and premature control valve replacements

## Opportunity Incumbents

- [Aspen DMC3](/Products/Aspen_DMC3) — Tool
- [Honeywell Profit Controller](/Products/Honeywell_Profit_Controller) — Tool
- [Emerson DeltaV Insight](/Products/Emerson_DeltaV_Insight) — Tool
- [Control Station PlantESP](/Products/Control_Station_PlantESP) — Tool
- [Manual Loop Tuning](/Products/Manual_Loop_Tuning) — DIY
- [Custom Excel Models](/Products/Custom_Excel_Models) — Spreadsheet

## Opportunity Win Conditions

**Kill Thresholds**:
- Operator override rate > 10% within 30 days of deployment
- Deployment time > 21 days per facility
- Valve oscillation reduction < 5% after 60 days of active use
- Pilot-to-paid conversion < 30% at the $300k annual contract tier
**Leading Metrics**:
- Time-to-first-loop-automated in hours
- Percentage of plant control loops operating in dynamic deadband mode
- Weekly operator override frequency per 100 loops
- Percentage reduction in aggregate valve travel volume
- Daily active usage rate among site process control engineers
**What Proves Right**: Process control engineers activate dynamic deadband adjustments on at least 15% of active loops within 30 days of deployment. Facilities sign $300k annual contracts because the software demonstrably reduces control valve oscillation and extends hardware lifespan. User cohorts maintain a 40% daily active usage rate to monitor loop diagnostics without reverting to manual tuning.
**What Proves Wrong**: Refinery operators routinely override the dynamic deadbands and force control loops back into manual mode during plant upsets. The software requires more than three weeks of custom engineering time per facility to map historical historian tags, destroying deployment economics. The adjustment logic fails to yield a quantifiable reduction in valve travel, making the software cost indefensible against legacy manual tuning.

## Opportunity Build Profile

**Hardest Part**: Writing safely to legacy Distributed Control Systems like Honeywell or Yokogawa without inducing runaway valve oscillation or violating strict plant safety constraints.
**Min Viable Scope**: Deliver an open-loop advisory model for a single non-critical process unit like a secondary cooling loop. Deliberately exclude closed-loop write-back execution, whole-plant optimization, and multi-variable model predictive control.
**Cold Start Problem**: Refineries refuse to grant closed-loop write access to unproven software. Break this by running strictly in read-only advisory mode, calculating the missed yield from static deadbands until the plant engineers trust the system enough to close the loop.
**Time To First Value**: 3-4 months. The gating step is passing the plant safety committee review and completing the mandatory open-loop shadowing period.
**Data Moat Available**: true
**Technical Difficulty**: High

## Neighborhood

### Incumbent in

- [Emerson DeltaV InSight](/Products/Emerson_DeltaV_InSight) — incumbent in · Products
- [AspenTech Aspen DMC3](/Products/AspenTech_Aspen_DMC3) — incumbent in · Products
- [Custom Excel Models](/Products/Custom_Excel_Models) — incumbent in · Products
- [Manual Loop Tuning](/Products/Manual_Loop_Tuning) — incumbent in · Products
- [Control Station PlantESP](/Products/Control_Station_PlantESP) — incumbent in · Products
- [Honeywell Profit Controller](/Products/Honeywell_Profit_Controller) — incumbent in · Products

### Applies thesis

- [Oil Refinery](/CompanyTypes/Oil_Refinery) — applies thesis · CompanyTypes

### Embodies

- [Software](/Theses/Software) — embodies · Theses

### Similar Opportunities

- [Autonomous Loop Tuning](/Opportunities/Autonomous_Loop_Tuning) — similar · Opportunities
- [Refiner Tuning Agent](/Opportunities/Refiner_Tuning_Agent) — similar · Opportunities
- [Feedstock Blending Optimizer](/Opportunities/Feedstock_Blending_Optimizer) — similar · Opportunities
- [Distillation Yield Engine](/Opportunities/Distillation_Yield_Engine) — similar · Opportunities
- [Algorithmic Refiner Energy Optimization](/Opportunities/Algorithmic_Refiner_Energy_Optimization) — similar · Opportunities
- [Refiner Energy Optimizer](/Opportunities/Refiner_Energy_Optimizer) — similar · Opportunities
- [Cracking Unit Diagnostics](/Opportunities/Cracking_Unit_Diagnostics) — similar · Opportunities
- [Freeness Control Agent](/Opportunities/Freeness_Control_Agent) — similar · Opportunities
- [Beneficiation Yield Optimizer](/Opportunities/Beneficiation_Yield_Optimizer) — similar · Opportunities
- [Ore Beneficiation Controller](/Opportunities/Ore_Beneficiation_Controller) — similar · Opportunities
- [AI Yield Optimization for Oleochemical Plants](/Opportunities/AI_Yield_Optimization_for_Oleochemical_Plants) — similar · Opportunities
- [Yield Setpoint API](/Opportunities/Yield_Setpoint_API) — similar · Opportunities
- [Combustion Optimization API](/Opportunities/Combustion_Optimization_API) — similar · Opportunities
- [Thermal Energy Telemetry for Mills](/Opportunities/Thermal_Energy_Telemetry_for_Mills) — similar · Opportunities
- [Generative Shift Handovers for Refineries](/Opportunities/Generative_Shift_Handovers_for_Refineries) — similar · Opportunities
- [Refiner Tuning Agent](/CompanyTypes/BCTMP_Mills/Opportunities/Refiner_Tuning_Agent) — similar · Opportunities
- [Alarm Flood Triage](/Opportunities/Alarm_Flood_Triage) — similar · Opportunities
- [Lift Optimization API](/Industries/Oil_and_Gas_Extraction/Opportunities/Lift_Optimization_API) — similar · Opportunities
- [AI Scale-Up for Chemical Producers](/Opportunities/AI_Scale-Up_for_Chemical_Producers) — similar · Opportunities
- [Bleach Dosing Engine](/CompanyTypes/BCTMP_Mills/Opportunities/Bleach_Dosing_Engine) — similar · Opportunities