recoverable heat and shifts energy demand back to utilities. As a result, a refinery may remain pinch-compliant on paper while steadily losing thermodynamic performance in operation. The distinction between pinch design intent and operating reality is summarised in Table 1 . Evidence from operating refineries Operating data from a 7 MMTPA medium-to- high-complexity Indian refinery were analysed over several years following a pinch-based revamp. Similar patterns have since been observed across multiple Indian refineries, indicating a systemic rather than site-specific behaviour. Within approximately three years of operation: • Heat recovery declined to ~78% of pinch design intent. • Fired-heater duty increased by ~5-6% at constant throughput. • SEC deteriorated by ~3-4%. • CO₂ intensity increased by ~4 kg/bbl (≈60 ktCO₂/year). “ Live measurements of temperatures, flows, and pressures are reconciled to remove sensor bias and noise, ensuring thermodynamic consistency ” All values are indicative and derived from reconciled operating data benchmarked against pinch design intent and normalised for throughput and crude severity. At prevailing fuel gas prices, this translated into an estimated incremental energy cost of INR30-45 crore per year, despite no equipment failures or immediate operability issues. Throughout this period, units continued to meet throughput and product-quality targets. Operators compensated instinctively – by increasing heater firing or steam extraction – allowing higher energy intensity to become normalised in routine operation. Over time, the original pinch intent faded from operational awareness. Why conventional monitoring misses the problem Most refineries monitor energy performance
using furnace efficiency, steam-to-feed ratios, or unit-level key performance indicators (KPIs). While necessary, these indicators do not explicitly reference pinch design intent or quantify system-level heat recovery degradation. Periodic energy audits can identify losses, but they are typically infrequent, retrospective, and resource-intensive. Between audit cycles, gradual heat integration degradation often becomes embedded in operating practice, remaining invisible until fuel costs or utility constraints become limiting. Digital energy management and heat integration digital twin Removing the ‘black box’ To close the gap between design intent and operation, the refinery implemented a Digital Energy Management System (real-time) centred on a heat integration digital twin. In this context, a digital twin refers to a continuously updated, steady-state thermodynamic model reconciled with live plant operating data. The underlying engine is based on physics- based thermodynamic mass and energy balances across the heat exchanger network. Live measurements of temperatures, flows, and pressures are reconciled to remove sensor bias and noise, ensuring thermodynamic consistency. From this reconciled data set, the model isolates effective exchanger performance parameters, such as overall heat-transfer coefficients (U-values) and approach temperatures, at an individual exchanger level. This enables the digital twin to distinguish between fouling-driven degradation, which manifests as gradual declines in effective U-values, and operational issues, such as bypassing or conservative control actions, which appear as step changes in approach temperature. By explaining why heat recovery has degraded, rather than merely indicating that it has, the model eliminates the perception of a ‘black box’. The system integrates live distributed control systems (DCS) data to enable heat integration performance tracking. From model to decision-support system Unlike conventional energy studies, which are periodic and retrospective, the digital twin
Refining India
36
Powered by FlippingBook