**4. Discussion**

The results of numerical simulations for the considered cases (Figure 3, Section 2.2) are compared in Figure 10. The value of change of average load temperature (ΔALT), calculated as the difference between ALT after 2 h of the process and the initial temperature of the system (i.e., 20 ◦C), is proportional to the amount of energy transferred to the load. ΔALT for case 1 serves as the reference value.

**Figure 10.** Load temperature difference (LTD) and change of average load temperature from initial temperature (ΔALT) after 2 h of the heat treatment (CPO—furnace corner power output).

Conducted CFD calculations show that equalization of the power output in each of the corners has a positive effect on the load temperature difference, especially in case of increased air–fuel equivalence ratio for natural gas burners and the resulting reduction of exhaust gas temperature. Replacing natural gas with biogenic syngas in the tested co-firing setup proved to achieve satisfactory efficiency of the heat treatment process—in four out of six analyzed co-firing scenarios the amount of heat transferred to the load exceeds 95% of the heat delivered to the load in the reference case (case 1). This effect is correlated with the lower SG flame temperature (Figure 6), and the fact that the streams of hot flue gases do not directly hit the load.

Analysis of the simulations results revealed a promising way of substituting 40% of natural gas in the preheating furnace with the considered biomass-derived renewable fuel, which leads to significant reduction of CO2 emissions, thus, smaller environmental impact. This is especially prominent in case 7, in which natural gas burners operate at λNG = 2.0 on two power levels: 72 kW (the ones near the syngas burners) and 360 kW (the other two burners), so in each of the furnace corners 360 kW of heat is generated. An important advantage of case 7 over cases 9 and 6 (especially the latter one) is that lower air-to-fuel ratio at the NG burners prevents the creation of strong stream of flue gas flowing out of natural gas burners operating at 360 kW (Figure A6), that could potentially disturb the combustion in the SG burners, which are inclined to less stable operation. Aggravation of this effect, caused by enhancing the stream of flue gas leaving the NG burners can also be noticed by comparing the gas flow patterns available in the Appendix A for the following sequences of cases: 1-3-2, 7-9-6, and 5-8-4 (Figures A1–A9).

Simultaneously, the temperature results show that the co-firing method assuming low air-to-fuel equivalence ratio and uneven distribution of power among the furnace corners (corner without SG burner: 216 kW, corner with SG burner: 504 kW) can still lead to satisfactory results (case 5). Analysis of the streamlines and temperature profiles for cases 5 and 7 shows that power equalizing on the one hand reduces the risk of local load overheating (Figure A5), but on the other hand promotes creation of strong jet-like flows, which hit and locally heat up inner walls of the furnace (Figure A7). It is likely that the optimal power balance between the furnace corners, where the scale of both phenomena is reduced and the efficiency achieves its maximum, lies between the analyzed values.

Based on these results, further work to evaluate the possibility of partial replacing natural gas consumption with alternative fuels (e.g., biomass-derived gaseous fuels or off-gases) can be done, especially for other types of furnaces (e.g., the ones continuously heating and melting stream of material). It is possible that in some cases, depending on the limits on LTD values, it may be necessary to increase the air–fuel equivalence ratio for syngas burners in order to lower SG flame temperature and improve thermal uniformity of the load.

**Author Contributions:** Conceptualization, P.J., A.K. and J.H.; methodology, P.J.; validation, P.J.; formal analysis, P.J.; investigation, P.J.; writing—original draft preparation, P.J.; writing—review and editing, P.J., J.H., A.K., K.B. and D.O.; visualization, P.J.; supervision, J.H.; project administration, J.H.; All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by European Union's Horizon 2020 research and innovation programme, grant number 723803. The APC was funded by the Institute of Power Engineering.

**Acknowledgments:** The authors would like to thank Jernej Mele (CPPE d.o.o.) and Matej Drobne (Valji d.o.o.) for support and providing process data.

**Conflicts of Interest:** The authors declare no conflict of interest.
