The operation of a pilot-scale online corrosion sensor system was studied at VTT's 100 kW grate pilot plant. The feedstock composition in tests was varied from 100% wood chips to a blend that also contained 40 en-% d.b. corn stover. The mass flow of alkali chlorides was varied with sulfur-containing additives. The measurements included electrical resistance (ER) of deposit and linear polarization resistance (LPR) for both St45.8 and AISI347H alloys. The number of fine particles (online data), alkali chloride mass flow, chloride content of the deposit, and flue gas composition (online data) were simultaneously measured to study how the furnace conditions correlate with the ER and LPR signals. Information on the risky furnace conditions can be obtained with these methods. The decrease in sensor signals after starting to feed in fuel blended with corn stover indicated increasing corrosion. The ER and LPR values started to increase again after changing the fuel back to 100% wood chips. The deposit composition, alkali chloride flow, and corrosion sensor signals predicted a low corrosion rate on the alloys with the corn stover-containing blend, when the mass flow of additive was sufficient to destroy the corrosive alkali chlorides before meeting the probe.