Development of novel fertilizer manufacturing processes using human urine

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Over 20 years of research on how to integrate and normalize urine as a resource has been conducted, including topics such as source-separation, fertilizer effectivity, public perception, and logistics. Another important aspect is the urine treatment method and how resources are recovered. This is because urine is 97% water, making the logistics of using urine as a fertilizer challenging. In addition, without treatment, the major nitrogen source in urine (urea) breaks down to ammonia and is lost to the atmosphere. This thesis investigated the technical feasibility of using different membrane and freezing techniques to concentrate and recover nutrients from human urine. It assessed different stabilization methods (to prevent urea breakdown), pre-treatment techniques, and process configurations to produce fertilizer products with different compositions. To prevent urea degradation and subsequent ammonia volatilization during the collection phase, urine was ‘stabilized' by adding an acid (citric acid) or a base (Ca(OH)2). The addition of Ca(OH)2 resulted in a saturated solution which would scale the RO membrane during concentration. Paper 1 and Paper 2 investigated two pre-treatment methods and showed that chemical addition (NaHCO3) and air bubbling can remove 85-98% of the excess calcium ions from the solution, thus significantly reducing potential RO membrane scaling during concentration. The calcium concentration in urine after stabilization varies based on urine composition. To minimize the addition of unwanted sodium ions, an equimolar dose of NaHCO3 is required. In Paper 1 it was determined that this dose (± 5%) could be determined by using conductivity as a proxy for calcium concentrations to help determine the required chemical dosage. In Paper 2, a model was developed to better understand the mechanisms of CO2 dissolution and CaCO3 precipitation. This model was used to optimize the time required to precipitate the maximum CaCO3 as well as the energy required to operate the air blower. In Paper 3, acid and base-stabilized urine were both concentrated using RO, and the two stabilization methods were compared. It was determined that membrane processes are not ideal for urine stabilized with an acid. This was due to the potential crystallization of uric acid dihydrate crystals which resulted in membrane scaling during concentration. Both pre-treatment methods for Ca(OH)2 stabilized urine were found to be equally effective at preventing membrane scaling and significantly improved flux and ion rejection. Air bubbling was chosen as the preferred pre-treatment method as it does not add additional ions (Na+ ), it reduces the urine pH to within the operating range of most common RO membranes (2-11), and it sequesters CO2 from the atmosphere. Reverse osmosis was successfully used to remove 60% of the water from real urine stabilized with Ca(OH)2 and pre-treated with air bubbling. The process recovered 85.5% of the urea and 98% of the potassium in the brine stream. Water removal was further improved to 70% with 79.5% of the urea and 98% of the potassium being recovered in the brine stream. This process produced a fertilizer with an N content of 1.9% and a K content of 0.5%. While RO is effective at concentrating stabilized urine, it also concentrates undesirable salts and pharmaceuticals together with desired fertilizer components. In Paper 4, the feasibility of nanofiltration (NF) as a pre-treatment to remove pharmaceuticals and salts was thus investigated. Two types of NF membranes were tested, a loose NF and a tight NF membrane. The NF permeate was then further concentrated with seawater RO membranes as described in Paper 3. A hybrid loose NF-RO configuration could remove 80% of water, more than 70% of the pharmaceuticals, 78% of the organics, and 44% of the total ions, however, urea recovery was only 56%. A tight NF-RO configuration could remove 80% of water, 90% of the organics, more than 99% of the pharmaceuticals, and 66% of the total ions, however, urea recovery was only 32.8%. Based on an economic analysis it is unlikely that the increased value of the product (due to increased purity) outweighs the additional cost of this pretreatment step. In Paper 5, eutectic freeze concentration (EFC) was investigated to further concentrate the RO brine (70% water removal), whilst simultaneously crystallizing undesirable salts, as this treatment method is not affected by membrane scaling. It was experimentally shown that at eutectic conditions, Na2SO4∙10H2O crystallizes simultaneously with ice. A theoretical mass balance of the RO-EFC process, including ice washing and recycle streams, showed that 77% of the urea and 96% of the potassium could be recovered with a 95% water removal. Over 98% of the phosphorus would be recovered as calcium phosphate during the urine stabilization step. The final liquid fertilizer would have a composition of 11.5% N and 3.5% K, and 3.5 kg of Na2SO4∙10H2O would theoretically be recovered from 1000 kg of urine. This research was the first to experimentally show that EFC can be used to concentrate human urine whilst simultaneously crystallizing salts. A high-level economic analysis showed that RO treatment processes have the lowest energy requirements (16 kWh m-3 , 70% water removal), followed by freeze concentration (119-162 kWh m-3 , 70-95% water removal), and lastly evaporative processes (154 -198 kWh m-3 , >95% water removal) which required the most energy. The fertilizer produced can either be sold as a niche (home gardening) or bulk (large agricultural) fertilizer. The size of the niche fertilizer market is important when determining a preferred treatment method. Assuming a feed supply of 7.5m3 urine per week, for a market size where only 0.14 m3 per week of niche fertilizer can be sold, RO-EFC produced the product with the highest value at R73 000. Alternatively, if the market size was 2 m3 per week, RO had the highest value product at R304 000. Treatment methods that produce a product with a higher nutrient content are preferred as bulk fertilizers. When selling fertilizers in bulk high density nutrient content is important to reduce transportation costs. For example, transport of the RO fertilizer, 75 km to farmland, would account for 3.2% of the gross fertilizer value whilst only 0.8% of the fertilizer value for the RO-EFC. The feed volume of urine required to make sufficient fertilizer for a small 20 ha wine farm using the alkaline dehydration treatment is 145 m3 which would take 19 weeks to collect from 8 shopping centers. This bulk fertilizer has a value of R76 000 compared to 0.14 m3 of niche RO-EFC fertilizer which has a potential value of R73 000. It would take less than a week to collect enough urine to produce the niche RO-EFC fertilizer. At this stage, focusing on the niche fertilizer market would be more profitable. It also indicates that urine collection and treatment need to become more mainstream and normalized before significant replacement of commercial synthetic fertilizers can be achieved. A membrane process with a variety of configurations can be used to concentrate human urine to produce a liquid fertilizer. Each configuration produces a product with a different composition and a commercial fertilizer with a comparable composition for each product could be found at a local garden center in Cape Town. However, the preferred treatment choice will be dependent on several factors such as process CAPEX and OPEX, fertilizer intended use (ornamental plants versus edible crops), the market size, the associated fertilizer regulations, and the scale at which urine collection is conducted. Overall, this work has shown that membrane processes can be used to concentrate stabilized urine whilst still achieving high urea (>79.5%) recovery and water removal (70%). This allows for significant scalability of urine treatment processes using RO membranes as this technology is already widely used to treat both brackish water and sea water at varying scales. However, not all urine stabilization methods are suitable for use before RO concentration. This research was the first to determine that acid stabilization results in the crystallization of uric acid dihydrate which would scale RO membranes and reduce efficiency. This research was also the first to show that EFC can be used to simultaneously crystallize ice and salts from real human urine that has been pre-concentrated with RO (70% water removal). The work also demonstrates how hybrid configurations, combining various urine concentration methods, can be used to produce products with different compositions and urea purity. Ultimately, this novel research makes a valuable contribution to the growing field of urine treatment and resource recovery.