Peak area variability in DMSO analysis

Hello,
I am running an impurity analysis method for DMSO with the following parameters:

FID Temperature: 250 °C
Injector Temperature: 250 °C
Column: Agilent DB-624
Flow: 5 mL/min (Hydrogen)
Initial temp: 100 °C, hold for 2 min
Ramp from 100 to 200 °C at 10 °C/min
Final temp: 200 °C, hold for 3 min
Injection volume: 0.5 microliters
Injection syringe volume: 1 microliter (auto-injector)
Split: 1:10
Liner: Ultrainert 5190-2295
FID flows:
• Hydrogen: 35 mL/min
• Air: 400 mL/min
• Nitrogen: 35 mL/min

I’m having serious issues with area repeatability when injecting the same vial 6 times in a row. I’m getting %RSD values around 9%, and my requirement is ≤ 5.0%
I’ve tried all kinds of adjustments—changing injection speeds, adding a 3-second viscosity delay… but nothing seems to solve the issue.

Can you help me?

Thank you very much.