Resolving Power Quality Challenges in a Dual-Source 33 kV API Facility

Facility: API Manufacturing Unit
Connected Load: 18 MW
Peak Demand: ~11.5 MW
Voltage Level: 33 kV / 415 V
Distribution Network: Dual-source 33 kV input, 2 x 1600 kVA transformers, 3 main LT panels, 12 load centres
Major Loads: HVAC, chillers, reactors, centrifuges, cleanroom VFDs, lighting, packaging lines

Load Flow Analysis

Objective: Evaluate system performance during peak and normal operation.

Simulated Scenarios:

• Case A: All loads running, including HVAC (Summer)
• Case B: Critical lines only, HVAC at 50% load (Monsoon)
• Case C: Expansion scenario (10% load increase)

Findings:

• Voltage at LCP-5 dropped to 393 V (6.6% below nominal) during peak in Case A.
• Bus-2 transformer load exceeded 94% utilization, triggering thermal alarms.
• Load imbalance between phases reached up to 17%, primarily on the cleanroom loads due to single-phase machinery.

Conclusion: Without optimization, the system would not support expansion beyond 10% load increase.

Harmonic Distortion Assessment

Measurement Points:

• Panel LCP-2 (serving HVAC VFDs)
• LCP-6 (serving packaging line)
• Main LT Panel Bus-1

THD Levels (Current):

• LCP-2: 8.2% (Exceeded IEEE 519 limit of 5%)
• LCP-6: 5.7%
• Main Bus: 4.6%

Analysis:

• Source: HVAC compressors using outdated VFDs with no harmonic mitigation.
• Ripple currents caused maloperation of PLCs in packaging lines (3 instances logged in 4 weeks).
• Measured K-factor of transformers: exceeded 9. Transformers were not K-rated, increasing winding heating.

Action Suggested:

• Passive filter bank at LCP-2
• K-rated isolation transformer addition to LCP-6

Transformer Loading & Diversity Analysis

Observations:

• T1 (1600 kVA) handled 82% of load while T2 remained at 42% due to misconfigured load transfer logic.
• Night shifts (when HVAC load drops) still showed T1 at 70% due to fixed feeder routing.

Simulation Result:

• Load balancing improved voltage stability by +1.8% at downstream feeders.
• Lowered transformer hotspot temperature by 8°C after reconfiguration.

Protection Coordination & Relay Discrimination

Tools Used: ETAP 21.0
System Inputs: Breaker curves, relay settings, manufacturer datasheets, transformer impedance

Findings:

• Breaker at Packaging Line Sub-panel was rated at 10 kA, while fault current was calculated at 12.3 kA RMS – > Under-rated breaker = safety risk.
• Upstream-downstream relay curves overlapped (no time selectivity), risking full bus shutdown on local faults.

Coordination Achieved:

• Primary relays (Main Incomer) set to trip at 0.6 sec
• Downstream relays adjusted to 0.2–0.3 sec zones
• Verified minimum clearance of 0.3 sec for 3 zones

Simulation Validation:

• All faults isolated within 1 level with no upstream tripping in test model
• System achieved 100% selectivity for 95% of fault conditions

Short Circuit Analysis

Key Metrics:

• Calculated fault at Main Bus: 28.7 kA (3ϕ, symmetrical)
• Bus-4 fault: 14.2 kA
• Breaker Capacity Check: 10 feeders found with <15% margin

Conclusion: Breaker upgrades recommended for 4 feeders and 2 capacitor bank panels.