Department Environmental Chemistry

Separate collection and treatment of urine optimizes nutrient recovery and enhances micropollutant removal from municipal wastewater. One typical urine treatment train includes nutrient recovery in three biological processes: anaerobic storage, followed by aerobic organics degradation concurrently with nitrification. These are usually followed by activated carbon adsorption to remove micropollutants. However, removing micropollutants prior to nitrification would protect nitrifiers from potential inhibition by pharmaceuticals. In addition, combining simplified biological treatment with activated carbon adsorption could offer a cheap and robust process for removing micropollutants where nutrient recovery is not the first priority, as a partial loss of ammonia occurs without nitrification. In this study, we investigated whether activated carbon adsorption could also take place between the three biological treatment steps. We tested the effectiveness of micropollutant removal with activated carbon after each biological treatment step by conducting experiments with anaerobically stored urine, organics-depleted urine, and nitrified urine. The urine solutions were spiked with 19 pharmaceuticals: amisulpride, atenolol, atenolol acid, candesartan, carbamazepine, citalopram, clarithromycin, darunavir, diclofenac, emtricitabine, fexofenadine, hydrochlorothiazide, irbesartan, lidocaine, metoprolol, N₄-acetylsulfamethoxazole, sulfamethoxazole, trimethoprim, venlafaxine, and two artificial sweeteners, acesulfame and sucralose. Batch experiments were conducted with powdered activated carbon (PAC) to determine how much activated carbon achieve which degree of micropollutant removal and how organics, pH, and speciation change from ammonium to nitrate influence adsorption. Micropollutant removal was also tested in granular activated carbon (GAC) columns, which is the preferred technology for micropollutant removal from urine. The carbon usage rates (CUR) per person were lower for all urine solutions than for municipal wastewater. The results showed that organics depletion would be needed when micropollutant removal was the sole aim of urine treatment, as the degradation of easily biodegradable organics prevented clogging of GAC columns. However, CUR did hardly improve with organics-depleted urine compared to stored urine. The lowest CUR was achieved with nitrified urine. This resulted from the additional organics removal during nitrification and not the lower pH or the partial conversion of ammonium to nitrate. In addition, we showed that the relative pharmaceutical removal in all solutions was independent of the initial pharmaceutical concentration unless the background organics matrix changed considerably. We conclude that removal of micropollutants in GAC columns from organics-depleted urine can be performed without clogging, but with the drawback of a higher carbon usage compared to removal from nitrified urine.

Adsorption on activated carbon is a common process to remove pharmaceuticals in wastewater treatment. Activated carbon adsorption is usually applied to wastewater with a low content of biological degradable organics, i.e. after biological treatment. Especially low molecular weight (LMW) compounds are known to compete with pharmaceuticals for adsorption sites. The goal of this study was to test the hypothesis that biological treatment is necessary for efficient pharmaceutical removal. Source-separated urine after anaerobic storage (anaerobically stored urine) and after aerobic biological removal of organics without nitrification (organics-depleted urine) were used in this study. In anaerobically stored urine 60% of the organic compounds were LMW organics, of which about 40% were acetate and propionate. 74% of the DOC and 100% of acetate and propionate were removed during aerobic biological treatment. To investigate the effect of the organic compounds on pharmaceutical removal, sorption experiments with 19 spiked pharmaceuticals and one artificial sweetener were conducted with powdered activated carbon. Ethanol, another LMW organic, was included in the study, as it is regularly used for pharmaceutical spiking thereby strongly increasing the DOC content. The experiments showed that the adsorption of the pharmaceuticals and the sweetener were hardly affected by the easily biodegradable LMW organics or ethanol. Therefore, it was concluded that biological pre-treatment is not necessary for efficient pharmaceutical adsorption. Since acetate, propionate and ethanol contribute substantially to the DOC content but do not absorb UV light, the latter is recommended as indicator for pharmaceutical removal in solutions with high contents of biodegradable LMW organics.

Urine has great potential to be an effective fertilizer due to its high nutrient content, however, it can contain potentially worrying pharmaceuticals. Our objective was to study whether urine storage and aerobic biological treatment, i.e. nitrification, was sufficient to remove pharmaceuticals or an additional treatment with activated carbon was necessary to produce a fertilizer from urine. We investigated the abatement of twelve pharmaceuticals, including antibiotics and antivirals, in laboratory experiments representing the treatment steps of anaerobic storage of source-separated human urine, stabilization using partial and full nitrification under acclimatized and non-acclimatized conditions, and treatment of nitrified urine using powdered activated carbon (PAC). Two-month-long-term storage of urine was insufficient to substantially degrade the pharmaceuticals, except for hydrochlorothiazide (>90%). In the partial and full nitrification fed-batch reactors, atazanavir, ritonavir, and clarithromycin were rapidly removed, with biotransformation rate constants greater than 10 L g_SS⁻¹d⁻¹. Darunavir, emtricitabine, trimethoprim, N4-acetylsulfamethoxazole, sulfamethoxazole, atenolol, diclofenac, and hydrochlorothiazide were degraded slowly, with biotransformation rate constants of < 1 L g_SS⁻¹d⁻¹. With 200 mg PAC L⁻¹, at least 90% of each investigated pharmaceutical was removed. Yeast estrogen screen tests and bioluminescence inhibition tests revealed efficient removal of estrogenicity (99%) and toxicity (56%) using nitrification, and a reduction of 89% and 64%, respectively, using 200 mg PAC L⁻¹. With our study, we provide biotransformation rate constants of compounds never previously investigated. We also show that a combination of nitrification and PAC adsorption enables the production of a safe fertilizer with sufficiently low pharmaceutical concentrations and no removal of beneficial nutrients.

Nitrification and distillation of urine allow for the recovery of all nutrients in a highly concentrated fertilizer solution. However, pharmaceuticals excreted with urine are only partially removed during these two process steps. For a sustainable and safe application, more extensive removal of pharmaceuticals is necessary. To enhance the pharmaceutical removal, which is already occurring during urine storage, nitrification and distillation, an adsorption column with granular activated carbon (GAC) can be included in the treatment train. We executed a pilot-scale study to investigate the adsorption of eleven indicator pharmaceuticals on GAC. During 74 days, we treated roughly 1000 L of pre-filtered and nitrified urine spiked with pharmaceuticals in two flow-through GAC columns filled with different grain sizes. We compared the performance of these columns by calculating the number of treated bed volumes until breakthrough and carbon usage rates. The eleven spiked pharmaceuticals were candesartan, carbamazepine, clarithromycin, diclofenac, emtricitabine, hydrochlorothiazide, irbesartan, metoprolol, N₄-acetylsulfamethoxazole, sulfamethoxazole and trimethoprim. At the shortest empty bed contact time (EBCT) of 25 min, immediate breakthrough was observed in both columns shortly after the start of the experiments. Strong competition by natural organic material (NOM) could have caused the low pharmaceutical removal at the EBCT of 25 min. At EBCTs of 70, 92 and 115 min, more than 660 bed volumes could be treated until breakthrough in the column with fine GAC. The earliest breakthrough was observed for candesartan and clarithromycin. On coarse GAC, only half the number of bed volumes could be treated until breakthrough compared to fine GAC. The probable reason for the later breakthrough with fine GAC is the smaller intraparticle diffusive path length. DOC and UV absorbance measurements at 265 nm indicated that both parameters can be used as indicators for the breakthrough of pharmaceuticals. In contrast to pharmaceuticals and DOC, the nutrient compounds ammonium, nitrate, phosphate, potassium and sulfate were not removed significantly. A comparison with literature values suggests that the amount of GAC needed to remove pharmaceuticals from human excreta could be reduced by nearly two orders of magnitude, if urine were treated on site instead of being discharged and treated in a centralized wastewater treatment plant.