Hydrothermal carbonization (HTC) of digested dairy manure produces hydrochar and a nitrogen-rich post-liquid. This study evaluated a submerged tubular expanded polytetrafluoroethylene (ePTFE) gas-permeable membrane (GPM) system for ammonia recovery from post-HTC liquid derived from digested dairy manure, focusing on the effects of feed temperature, acid circulation rate, and feed volume-to-membrane surface area ratio (FV/MS). Compared with filtered digested manure, the post-HTC liquid had a similar total ammoniacal nitrogen (TAN) concentration (1151 vs. 1164 mg N L⁻¹) but a slightly higher pH (8.38 vs. 7.94), which favored ammonia transfer. Increasing feed temperature from 20 to 60 °C raised 24 h TAN recovery from 63.5% to 99.6% and average TAN transfer flux from 11.1 to 17.7 g m⁻² d⁻¹, although it also increased water vapor crossover and diluted the acid trapping solution. Increasing acid circulation from 10 to 30 mL min⁻¹ improved 48 h TAN recovery from 85.2% to 94.0%, while a further increase to 50 mL min⁻¹ produced only a small additional gain. In contrast, elevating FV/MS from 0.015 to 0.045 m³ m⁻² reduced 48 h TAN recovery from 94.0% to 59.6% while increasing average TAN transfer flux from 8.2 to 16.2 g m⁻² d⁻¹ because larger feed volumes sustained the concentration driving force for a longer period. Nitrogen mass balance showed that more than 94–99% of the initial TAN was accounted under most conditions, with only a small fraction lost to volatilization.

This study focused on the optimization of hydrothermal carbonization (HTC) of nutrient-rich dairy manure digestate using response surface methodology based on a face-centered central composite design (FCCD-RSM), aiming to simultaneously maximize hydrochar (HC) yield and nutrient recovery. FCCD-RSM was employed to systematically evaluate the impact of HTC reaction temperature (180–240 °C) and reaction time (20–90 min) on HC yield, phosphorus (P) content in HC, and total ammonia-nitrogen (TAN) concentration in the HTC liquid. Results showed that the P content in HC and the TAN concentration in the HTC liquid increased with increasing HTC temperature and reaction time, whereas HC yield decreased. Based on experimental data, quadratic polynomial models were developed. Using RSM with equal weighting assigned to the three responses, the optimized conditions were determined to be 216.5 °C and 84.1 min, resulting in an HC yield of 79.03 ± 2.61%, P content in HC of 1.76 ± 0.04%, and TAN concentration in the HTC liquid of 891.67 ± 16.67 mg/L. These values were close to the predicted values and fell within the 95% confidence intervals. Therefore, the optimized HTC parameters can simultaneously and effectively produce P-rich HC and TAN-rich liquid, providing insights into a more sustainable approach to circular waste management and nutrient recycling on dairy farms.