Ultrafiltration Model Development and Experimental Validation for Application to the Enhanced Actinide Removal Plant (EARP)

  • Kyriacos Hadjidemetriou

Student thesis: Phd


The Enhanced Actinide Removal Plant (EARP) at the Sellafield Site (UK) generates and concentrates ferric flocs with incorporated radioactive components produced from acidic feeds arising from upstream spent nuclear fuel reprocessing plants. The EARP process is based on the chemical precipitation (CP) of an iron (oxy)hydroxide floc (ferric floc) followed by ultrafiltration (UF) stages for waste volume minimisation. As the reprocessing plants shift to post operational clean-out (POCO) and decommissioning, the effluent chemistry received by EARP could potentially be changed. The current UF system deals with large volumes of waste effluent and has operated using set conditions for the past two decades. However, the generation of a model of the UF process can potentially provide improvements in the parametric conditions of the system operation and membrane filterability. In this industry-based PhD, experimental and modelling work have been undertaken. The aim of this work is to underpin future plant operation, by creating a predictive modelling capability to evaluate future plant performance under a range of different scenarios with a particular emphasis on UF membrane fouling. Initially, a 0D predictive mathematical UF model has been created with key parameters such as UF membrane porosity and filter cake thickness taken from experimental measurements. This model can predict the pressure loss in the membrane unit and the permeate flowrate with good accuracy (average error of 1.2 and 5.6% for the pressure loss across the membrane and permeate flow, respectively). The predictive capability of the UF model was assessed through UF experimental trials performed on a bespoke UF rig designed to replicate the function of the EARP UF system as closely as possible using water and ferric floc (ferrihydrite, 0.216 g l-1 iron) as the process fluids. The model was also able to predict the observed permeate flowrate for a range of inlet flow rates and inlet pressures. UF membrane performance is affected by the fouling deposition on the surface and the pores within the UF membrane. Therefore, a fouling resistance factor was incorporated into the predictive model to evaluate the UF behaviour in the presence of fouling deposition on the interior of the membrane. The development of the model has taken into consideration the use of different inlet parameters and the incorporation of a mathematically derived fouling resistance factor. The utility of the predictive UF model has also been extended to predict the permeate generation and the system performance while processing different ferric floc phases such as goethite. This extension of the model utility can guide future EARP operations of a potential ferric floc phase to be treated within the UF system as the chemistry of the system is altered during POCO activities. Finally, the thesis progresses to understand through experimental work the effect of the inclusion of the POCO decontamination agent, citric acid within a ferrihydrite based process fluid on the UF system performance of EARP. These experiments contribute in several ways to understanding of the operating conditions of UF system within EARP once the decontaminating agent citric acid is added in the iron waste effluents received by the system at the SL site.
Date of Award1 Aug 2022
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorThomas Rodgers (Supervisor) & Claudio Pereira Da Fonte (Supervisor)


  • Effluent Treatment
  • Nuclear Decommissioning
  • Fouling Model
  • Goethite
  • Ferrihydrite
  • Experimental
  • Modelling
  • Ultrafiltration
  • Enhanced Actinide Removal Plant
  • Membrane Fouling

Cite this