The combustion of phosphorus (P)-rich biomass has a significant potential to increase the stock of biomass resources available for renewable heat and power production. In addition, the P-rich ashes have the potential as a fertilizer and could reduce the need for non-renewable P resources in agriculture. However, several technical challenges must be resolved to realize this potential.

During combustion, the ash-forming matter in the fuel undergoes numerous chemical transformations, which can result in the formation of ash melts. Excessive melt formation can lead to durable ash deposits in and around the combustion zone and on heat exchangers, which can lower combustion performance and, in severe cases, lead to a complete shutdown of the process. Further, mono-combustion of P-rich residues such as sewage sludge results in the formation of phosphates with poor plant availability, which can significantly limit the value of the ash as a fertilizer. By co-combustion the sludge with K-rich biomass, it may be possible to alter the chemical speciation of P towards more plant-available phases while simultaneously managing the risk of ash-related operational issues. This work investigates the effect of combustion technology, fuel ash composition, and chemical association of P in the fuel on the fate, i.e., distribution and speciation, and ash transformations of P in combustion and co-combustion of biomass and sewage sludge.

The basis of the study was experiments performed in three different combustion technologies, including a fluidized bed (5 kW, 730-800 °C), a fixed bed (20 kW, 950-1250 °C), and a powder burner (150 kW, ~1100°C). The fuels and fuel mixtures included P-rich and P-poor woody biomass, agricultural residues, and sewage sludge, which constitute a wide range of ash compositions in terms of K, Ca, Mg, Fe, Al, Si, and P. The residual ashes from the experiments were collected and chemically characterized with the original fuels and fuel mixtures to determine the ash transformation reactions of P. The experiments were complemented by thermodynamic equilibrium calculations (TECs), which aided the interpretation of experimental data and predicted the risk for operational issues related to the melting of coarse ash fractions.

The major share of fuel P was found in coarse ash fractions such as bed ash particles, bottom ash, slag, cyclone ash, wind side deposits, and coarse fly ash. A low share of fuel P was found in fine ash fractions such as leeside deposits and PM1. This generally matched the predictions by TEC, which indicated that P was stable in condensed phases at the relevant compositions and conditions during the combustion experiments. The powder burner experiments produced the highest share of fuel P in PM1 (4-14 wt.%), followed by fixed bed combustion (<4 wt.%), with fluidized bed combustion having the lowest share (<0.6 wt.%). In addition, the experiments with sewage sludge indicated a significantly lower P share in PM1 for a given combustion technology than the other biomass fuels, ranging from <0.2 wt.% in the fluidized bed and <1.2 wt.% in the fixed bed.

Combustion and co-combustion of woody biomass and agricultural residues resulted in the formation of a wide range of ortho-, pyro-, and metaphosphates associated with K, Ca, and Mg. Combustion of woody biomass generally resulted in a high share of Ca-orthophosphates, whereas agricultural residues had a higher share of K-rich ortho- and pyrophosphates. Irrespective of biomass assortment, the speciation of P in the ash from combustion and co-combustion followed general trends with respect to the fuel ash composition of the biomass mixture. The frequency and share of pyro- and metaphosphates identified in the coarse ash fractions tended to increase with the relative concentration of P to K, Ca, and Mg in the fuel mixture. A similar correlation was found between the share of K-rich phosphates and the relative concentration of K to Ca and Mg.

The crystalline phosphate phases identified in the coarse ash fractions from sewage sludge and K-rich biomass experiments were mainly Fe-rich and Ca-rich orthophosphate. The frequency and share of Fe-rich orthophosphates decreased with the relative P to K, Ca, and Mg concentration in the fuel mixture. However, the sewage sludge mixtures were less prone to form K-rich orthophosphates than the biomass mixtures for a given composition in terms of P to K, Ca, and Mg.

Based on TECs, it was possible to qualitatively predict ash-related issues related to the melting behavior of coarse ash fractions, such as slag formation, for woody biomass and agricultural residues by considering the K, Ca, Mg, Si, and P content in the fuel. The share of network formers (SiO2, PO2.5) to total ash oxides had the largest overall influence on the melting tendency of the ash mixture, followed by the ratio of K2O to total network modifiers (K2O, CaO, MgO), which had a high impact on ash mixtures with high relative shares of SiO2. The slagging tendency of fuel mixtures with a high share of sewage sludge could not be predicted based on the melting behavior of the K-Ca-Mg-Si-P-O system due to the high relative share of Fe and Al. The experimental results indicated that the slagging tendency of the sewage sludge was significantly improved by co-combustion with moderate amounts of wheat straw or sunflower husk.

Based on the combined results, it was possible to establish four fuel ash molar ratios correlated with the speciation of P in the produced coarse ash fractions and the risk of slag formation in fixed-bed combustion. These ratios were used to recommend practical fuel mixing strategies that could enable the production of combustion ashes with high P-plant availability while simultaneously managing the risk of severe slag formation.