The application and performance of urine diversion and solids dehydration in the management of human waste in remote alpine environments Geoff Hill*, MSc Department of Geography University of British Columbia 1984 West Mall Vancouver, BC V6T1Z2 [email protected] Phone: (604) 822-2663 Cell: (604) 505-3656 Fax: (604) 822-6150 Greg Henry, PhD Department of Geography University of British Columbia 1984 West Mall Vancouver, BC V6T1Z2 [email protected] Phone: (604) 822-2663 Fax: (604) 822-6150 *Corresponding Author

Abstract In order to minimize the costs and impacts associated with human waste management in alpine environments, we designed and tested three alternative waste treatment systems at the Kain Hut, Bugaboo Provincial Park, BC, Canada. By quantifying the mass of excreta deposited per toilet use, we were able to compare the performance of urine-diversion (UD), solar-dehydration (UD12V), and 110V-evaporation (UD110V) against the standard all-in-one barrel collection (BFO). UD significantly reduced human excreta by 60% potentially saving $108 per barrel when removed from Bugaboo Park with a Bell 407 helicopter. UD110V consistently raised the air temperature and reduced the relative humidity, in an insulated basement chamber, dehydrating the excreta mass by a significant additional 34% beyond UD (94% less than BFO). More research needs be conducted on the optimal collection container for the resulting condensed solids and on the ecological impacts of locally discharged urine. Introduction Human waste management at remote alpine sites is very challenging. Remote sites frequently lack standard municipal infrastructure including road access, sewerage, electricity, and water supply. Without these basic services the removal or treatment of human waste can become an expensive, intensive, and offensive task. Additional challenges include: climate, intense UV light, short summers, large diurnal fluctuations, frequent freeze-thaw events, extreme weather, shallow weak soils, limited vegetation, and challenging terrain (Weissenbacher et al. 2008). Nonetheless, the proper management of human waste is essential in order to prevent environmental contamination, ensure adequate user sanitation, and meet legal requirements. European alpine sites experience greater usage than in North America, which has stimulated more stringent legal regulations and a wider selection of suitable waste treatment technologies (Becker et al. 2007). With increasing alpine usage and inadequate waste management systems in 25% of US National Parks (Climiburg et al. 2000), it is foreseeable that regulations may become more stringent which may in turn demand more comprehensive solutions. However, there is a dearth of applied research in the North American alpine on appropriate waste management systems, their costs, and associated environmental impacts. Becker et al. (2007) provide a summary of waste management solutions employed in Europe. Many European systems rely on electrical power and running water making them impractical at the majority of North American alpine sites, which rarely have electricity or running water. With the exception of a small number of private lodges, most alpine destinations do not have road access, site power, sewerage, or running water. Many sites with these constraints exist in National, Provincial, or State Parks, which in some cases are leased by clubs and associations such as the Alpine Club of Canada (ACC) or the American Alpine Club (AAC). Human waste management systems currently employed in these NA alpine sites include; i) no management; ii) single use collection and pack-out bags; iii) disposal down a steep drop such as a crevasse or rock face; iv) deposition into a pit that is covered with earth when full; v) deposition into a replaceable container that is removed by hand, quad, snowmobile, pack animal, or helicopter; vi) deposited into a container whose contents are incinerated with fuel after evaporating liquids; vii) treated onsite through aerobic decomposition (composting toilets).

The majority of human waste at alpine huts in the Canadian National Park and Provincial Park systems is flown out by helicopter (Devlin  &  Sparks  2010). The average cost per barrel, in 2010, at a site with 25 full barrels, including helicopter time, fuel, five Alpine Club of Canada staff, and a pump truck at the closest accessible road, was $180/barrel. At popular alpine huts, the total expense can be many thousands of dollars. With the objective of reducing the expense, hazard, and greenhouse gas emissions associated with helicopter transport of human waste, we chose to assess the performance of urine diversion (UD) and solid excrement dehydration in comparison with the standard barrel-fly-out system. UD has been shown to have both cost savings and environmental benefits at many scales of application. At the municipal scale UD alleviates the nitrogen load on waste water treatment systems, reducing energy, GHGs, and cost associated with denitrification (Shiskowski 2009). Captured and oxidized urea can be processed into agricultural fertilizer reducing dependence on fossil fuel derived ammonium nitrate (Höglund et al. 1998). Sridevi et al. (2009) and Heinonen-Tanski et al. (2007) show urine’s value as fertilizer for bananas, barley, and cucumbers. Lienert and Larsen (2006) describe the sociological challenges associated with UD toilet technology and optimistically report 86% acceptance at two public institutions in Switzerland. Of principal interest to alpine waste management, urine comprises up to 75% of a human’s daily excrement mass, contains minimal pathogens, and can fertilize local plant communities (Schouw et al. 2001). By eliminating urine from the helicopter-transported fraction, each barrel could contain up to four times as many toilet visits. However, there are numerous factors that could affect the actual performance of UD in the alpine, including user perception, education, gender / use ratio, child / adult ratio, and maintenance challenges. Dehydration of the remaining solids, post UD, could further reduce waste treatment mass, volume, cost, and handling hazards. High insolation, common at many alpine sites, could be used advantageously in the dehydration of human waste. In addition, some alpine sites have PV panel, micro-hydro or generator power, which could be used during off-peak hours to accelerate the dehydration of human waste. As far as we know, there have been no experimental comparisons in an alpine environment between the standard barrel-fly-out system and alternative waste treatment technologies. Methods We chose the Conrad Kain Hut, Bugaboo Provincial Park, BC, as a site to test three alternative waste treatment technologies. The hut sits 5km from and 700m above the trailhead, 45 km west of Brisco, BC. Accommodating 40 overnight occupants, it is used principally in the summer by hikers, climbers and guides. It is one of the more popular destinations in the Canadian alpine and is serviced with propane for cooking, a micro-hydro generator for heating and lighting, running water and plumbing for drinking and cooking, and a greywater disposal field in a natural sedge meadow. There are three outdoor toilets, one close to the hut and two down a short flight of stairs. Prior to our experimental manipulations the toilet close to the hut was used as a urine-only-toilet; a mesh grate dissuaded fecal matter additions. Urine from the urine only toilet was diverted into the greywater disposal pipe. The hut and toilets sit upon a small bedrock knoll with unobstructed solar exposure until mid afternoon when Snowpatch Spire interrupts direct incoming solar radiation.

We designed and assessed three alternative toilet waste management systems that could be retrofitted into any standard barrel-fly-out toilet (BFO). The simplest system was urinediversion (UD), which included both a men’s urinal and urine diverting seat from EcoVita (Bedford, MA) (Figure 1A and 1B). The second involved the urine-diversion system plus solar dehydration (UD12V). This system transfers incoming solar radiation to sensible heat inside a thin flat transparent panel; this hot air is then driven through ducting by a fan and PV panel to the surface of the excrement pile. The 0.5m^2 solar hot air panel, 100 CFM fan, and 5W PV panel are sold by Clear Dome Solar (San Diego, CA). We designed our own solar dehydrating toilet system based on Arnold’s design, a RMNP Ranger, as the only commercially available unit (SWSLOO) was too large to be retrofit into most existing outhouse structures (Arnold 2010) (Figure 1C). The third system combined urine-diversion with a 110V 800W heater and a 110V 110CFM blower and exhaust fan inside an insulated chamber (UD110V) (Figure 1D). The toilet closest to the Kain Hut was chosen for UD110V due to its proximity to 110V outlets. The basement chamber at this toilet was insulated with 4cm thick styrafoam boards. Data were collected during two sample periods, August 15-17 and September 3-5, during which time access to the other toilets was restricted so as to account for all toilet uses. BFO, UD, and UD 12V treatment systems were established at the lower two toilets for 3-6 day periods according the following schedule: BFO - BFO August 14-18, BFO - UD12V August 18-20, UD - UD12V September 4-10, BFO - UD September 14-19. During these sample periods access to the UD110V upper toilet was restricted as much as possible without creating line-ups. In addition, hut visitors were instructed to use all non-restricted toilets equally during their stay so as to ensure an even and unbiased distribution of toilet use. In order to determine mass reduction performance with respect to the standard BFO, we recorded the number of door counts at 6-24 hour intervals over the course of the 3-6 day sample periods. The interval and period length depended on the intensity of hut visitation; we increased sampling intensity with elevated visitation. We targeted 10-30 toilet uses per interval in order to maximize the number of intervals while minimizing differentiable mass change at the collection barrel below each toilet. Change in barrel mass was determined by weighing the collection barrel before and after each sampling interval with a veterinary pet scale. Door counters were EPC-MAG1 model made by Inter-Dimensional Technologies, Inc. (Hop Bottom, PA). A ten second delay function was employed in order to eliminate erroneous readings caused by wind or door closing errors. We subtracted the unit’s final count from its initial and divided the difference by two in order to obtain the total toilet uses. Dividing the change in barrel mass by toilet use eliminated the effect of variable sampling interval length and established a robust quantifiable baseline in the assessment of remote site waste treatment performance. A simple mass balance equation was used to quantify performance. Data loggers were used to collect ambient and treatment system air temperature and relative humidity data. All three alternative treatment systems were tested twice. BFO was tested three times. Combined there were nine treatment runs conducted between July and September 2010. Each run was divided into three to six sample periods. Measurements with less than 5 toilet uses per sampling period were not used in order to reduce variability. JMP 8 (SAS Institute) was used to analyze our data. The alpha value was set at 0.05. One outlier was removed from the BFO treatment data set after it was discovered that a

dysfunctional door latch caused an overestimation of toilet use. No other alterations or transformations were made.

A  

B  

D  

C  

ii  

i  

Fig. 1 Alternative toilet waste management treatments at Kain Hut, Bugaboo Park, BC. A) EcoVita Privy Kit 501 (optional child riser not used). B) EcoVita urinal with 1” braided drain pipe to collection barrel. C) Lower toilets with UD12V solar hot air panel (i) above a 5W PV panel (ii) and full excrement barrel-fly-out containers. D) Upper Kain Hut toilet insulated basement, 110V heater, and 110CFM exhaust fan (not visible) positioned around a 200L barrel.

Results The installation of the urine diversion seat and urinal required one hour (Figure 1A, 1B). The solar hot air system was tested prior to installation on August 16th on an exposed meadow adjacent to the Kain Hut. The sky was cloudless and winds were calm over the course of the day. The solar hot air panel consistently raised the air temperature and reduced the relative humidity for eight and a half hours by an average of +10oC and -14% with a maximum heating of +15oC and drying of -19%. Wind speeds at the outlet of the vent varied from 0-3 m/s. The solar hot air system required eight hours to plan and install at the lower toilet site (Figure 1C). Over the course of two sample periods, spanning four days, the treatment consistently raised the air temperature and reduced the relative humidity for 6.8 hours per day. The hot air panel produced a maximum of 3m/s air flow, heating of +7oC, and drying of -18%. UD110V system assembly and testing required ~15 days. During a representative 20 hour sample interval the system increased the average basement temperature and reduced the relative humidity by an average of +24.7oC and -44%, up to a maximum of +30.5 oC and 63%. The system averaged an actual temperature of 31 oC and 17% relative humidity.

Change in mass / use (kg/use)

Change in barrel mass per toilet use data were compared within treatment type with robust, rank sum, non-parametric Wilcox and Kruskal-Wallis tests; none of the treatment runs were significantly different. In order to increase sample sizes, we grouped treatment runs into treatment types (Figure 2). The relationship between mean change in excreta mass per toilet use by treatment type was significant (p<0.0001) with largest mass associated with BFO toilets (0.262 ± 0.073 kg/use), followed by UD (0.105 ± 0.33 kg/use), UD12V (0.082 ± 0.042 kg/use), and UD110V (0.016 ± 0.017 kg/use).

0.4

A  

B  

B  

C  

0.3

Treatment Type BFO UD UD12V UD110V

0.2 0.1

0 BFO

UD UD12V Treatment Type

UD110V

Figure 2. Change in excreta mass per toilet use by treatment type (kg/use). Mean and 95% confidence intervals denoted by wide central line and narrow upper and lower lines. KruskalWallis non-parametric chi-square approximation test on treatment type significant

(p<0.0001). Significantly different means denoted with different letters (A, B, C) as determined by Tukey HSD ANCOVA slope comparison (Figure 3). An ANCOVA was used to isolate the statistical significance by testing for a difference between treatment type regression slopes (Figure 3A). The ANCOVA model was generated in JMP 8 with three variables: treatment type, toilet use, and the interaction between the two variables. The RSquare Adjusted values of the change in mass (kg) by toilet use by treatment type were 0.97 for BFO and UD, 0.53 for UD12V and 0.48 for UD110V. The Actual by Predicted Plot show a strong fit, a high explanation of error (RSquare = 0.97), and statistical significance (p<0.0001) based on an analysis of variance (Figure 3B). The residual plot raises no concerns with regards to heteroscedasticity or treatment type bias (Figure 3C). There is a strong positive and significant (p<0.0001) relationship between toilet uses and the change in mass of the collection container (residuals not shown). As expected, each time the toilet is used, excrement mass is added to the collection container. A Tukey HSD was used to test for differences between treatment type slopes. Significantly different slopes were discovered; BFO and UD 110V were significantly different from each other and both UD and UD12V. The interaction term between toilet use and treatment type was also significant due to the UD110V treatment (p<0.0001) (Figure 3F).

A  

B  

C  

  Figure 3. Analysis of Covariance (ANCOVA) testing difference between treatment types. A) Regression by treatment type B) Plot of ACOVA model predictions vs. actual change in mass (RSq=0.97). C) Plot of ANCOVA predictions vs. actual residuals.

Discussion Schouw et al. (2002) found that the average mass of excreta generated per toilet use of urine, wet feces, and combined excreta equates to 0.167 kg, 0.153 kg, and 0.320 kg. In comparison, our mean urine, wet feces, and combined excreta masses per toilet use were 0.186 ± 0.079 kg (not shown) and 0.105 ± 0.033 kg and 0.262 ± 0.073 kg. The variability between Schouw et al. (2002) and our findings could be explained by many physiological, behavioral, or environmental parameters including age, sex, activity level, social norms, or altitude. However, given that our values are within 30% of Schouw et al.’s, noting that there are likely numerous behavioral and physiological differences between farmers and backcountry travelers, we can assume that our methods were appropriate. Our results indicate that with the addition of a urine diversion seat and urinal up to 0.157 ± 0.080 kg per use of excreta can be eliminated from the barrel-fly-out system (Equation 1C Equation 1D). This equates to a 60% reduction in barrel-fly-out mass. Equipped with urine diversion equipment, each barrel will hold 2.5 times as many excrement deposits as compared with standard all-in-one barrel collection systems, greatly reducing the total numbers of barrels filled at each site. A urine diversion seat and urinal costs less than $200CDN. Installation takes less than two hours. Sites without a preexisting greywater treatment system would need to construct an adequate dispersal and leach field. For sites that generate more than three barrels of excrement per year (the max load of a Bell 407), investment into a UD system could reduce the total cost of barrel removal from $180 to $72 per barrel. UD should incur minimal additional maintenance costs as long as educational signage is posted. With non-significant differences between UD and UD12V, we are unable to conclude whether solar dehydration is a viable waste reduction treatment (Figure 2). However, UD12V’s mean excreta mass per toilet use was lower than UD’s indicating that further research with greater statistical power may be all that is required to show significance (Figure 2). A large solar hot air panel or the positioning of the panel in an area with greater solar exposure, as was shown in the pre-installation trial, would likely increase the dehydration effect. In addition, we sampled near the end of the summer when incoming solar radiation was diminishing and cloud cover was increasing. The UD110V treatment had the lowest mean excreta mass per toilet use. During the sample period the elevated temperature and lowered relative humidity within the insulated basement created strong evaporation at the surface of the excrement pile. Between occasional sample periods the mass of water evaporated was greater than or equal to the mass of excreta added. We used these instances to estimate the mass of water evaporated per use (0.089 ± 0.037 kg). Multiplying the water evaporated per use by the average uses per hour during the sample intervals (2.1) we obtain an estimate of the evaporation rate per hour (0.187 ± 0.078 kg/hour). In comparison, we determined the average rate of evaporation from a 20L pail of water to be 0.166 ± 0.045 kg/hour. It is conceivable that the estimated rate of evaporation from feces is on par with the measured rate from a pail of water considering the wicking nature of toilet paper. We assumed the ratio of urination to defecation by toilet was constant; despite this toilet’s potential preference for night-time urination visits, it is the furthest toilet from the only trail leading to and from the hut, which likely balances any bias in excrement type ratio per use. Future studies could survey this ratio in order to eliminate this uncertainty.

In order to measure the mass of the UD110V barrel between each sampling interval, the insulated door had to be removed. This resulted in total heat loss to and humidity gain from the surrounding environment. In addition, the 110V circuit servicing the toilet experienced occasional interruptions due to higher priority loads at the hut. These inconsistencies through time created a weak R2 value of 0.48 when compared with BFO and UD’s strong R2 Adjusted values of 0.97 and 0.97. The microclimate surrounding the BFO and UD barrels was much more stable; diurnal temperature and humidity varied less than 10oC and 20% and were unaffected by sampling interruption. Conventionally excrement barrels are flown to the nearest road where they are met by a pump truck and vacuumed into a larger containment tank. The diluted sludge is then transported to the nearest sewage treatment plant. During this experiment solid excrement waste was collected in doubled large garbage bags, which were transported to the landfill along with the single use pack-out excrement bags. Trials need to be conducted to determine the optimal containment vessel and disposal strategy. Urine diverted from standard collection barrels during the experimental period was added to BFO waste because of the unknown impact of urine discharge into the small stream below the lower hut toilets. In the future all urine will be routed into the greywater discharge pipe used by the site’s earlier urine-only toilet. In addition it will be important to evaluate the ecological impact of urine discharge. If dispersed into the organic horizon of alpine soils we expect no harmful impacts to the local ecosystem or unpleasant odors, both of which would be of concern to National and Provincial Park authorities (Uder, Larsen, Gujer 2006). On the contrary, ample research shows that urea is readily oxidized and nitrified by soil bacteria into plant available nitrates (Heinonen-Tanski et al. 2007, Sridevi et al. 2009). Urine diversion may also improve the performance of remote site pit toilets by greatly reducing the volume of liquid leachate flowing through the pile of excrement. In addition, elevated soil microbial activity response due to fertilization may improve the decomposition of adjacent fecal waste and outcompete or consume harmful bacteria, protozoa, worms, and other transportable pathogenic microorganisms known to cause illness and disease in humans (Chapin et al. 1995, Wang et al. 2010) Conclusion By weighing the change in containment vessel mass per toilet use we developed an index for the quantification of remote waste treatment system performance. Urine diversion through urinals and urine diverting seats can easily and inexpensively reduce barrel-fly-out mass by 60%, potentially saving up to $108 per barrel when removed en mass with a Bell 407 helicopter from the Kain Hut, Bugaboo Park. While not statistically significant, the addition of a solar dehydration system may further reduce remote site waste mass an additional 9%. The addition of 110V heaters and fans to a urine diverted toilet significantly reduced barrelfly-out mass by 94% assuming all toilets were used equally for solid and liquid deposits, as was requested of site visitors and managed by regulating toilet access. While barrel-fly-out and urine diversion only treatments are consistent through time, the performance of solar and 110V dehydration depend on a number of variables including usage rate, available insulation, and consistency of 110V power. The practical application of urine diversion may be constrained by the removal and disposal of condensed solids, which were previously removed

and disposed of by a vacuum pump truck. In addition future research needs to determine the impact of urine diversion from barrel-fly-out toilets and pit toilets on local ecosystems and on visitor experience. Acknowledgements Mountain Equipment CO-OP, the Alpine Club of Canada, Backcountry Energy Environment Solutions (BEES), and the Natural Sciences and Engineering Research Council (NSERC). Knut Kitching dedicated the summer of 2010 to human waste management and deserves much thanks. Benard Faure and the ACC hut custodians provided considerable assistance in with data collection. In addition thanks to Dr. Sue Baldwin and Dr. Anthony Lau for their assistance in project design and Joe Arnold for the design of his solar dehydration toilets. References Arnold (2010, July) Solar Dehydrating Toilets in Rocky Mountain National Park. Paper presented at the Exit Strategies: Managing Human Waste in the Wild Conference, Golden, CO. Becker W, Schoen MA, Wett B (2007) Solar-thermic sewage sludge treatment in extreme alpine environments. Water Science & Technology, 56(11), 1-9. Chapin  FS,  Shaver  GR,  Giblin  AE,  Nakelhoffer  KJ,  Laundre  JA  (1995)  Responses  of  Arctic   Tundra  to  experimental  and  observed  changes  in  Climate.    Ecology,  76(3),  694-­‐ 711.     Climburg A, Monz C, Kehoe S (2000) Wildland Recreation and Human Waste: A Review of Problems, Practices, and Concerns. Environmental Management, 25(6), 587-698. Devlin  R  and  Sparks  S  (2010)  Blackcountry  Blackwater  Management  Options  Analysis.     Stantec  consulting  report  prepared  for  Backcountry  Energy  Environmental   Solutions  (BEES),  Edmonton,  Alberta.   Heinonen-Tanski H, Sjöblom A, Fabritius H, Karinen P (2007) Pure human urine is a good fertilizer for cucombers. Bioresource Technology 98, 214-217. Höglund C, Stenström TA, Jönsson H, Sundin A (1998) Evaluation of faecal contamination and microbial die-off in urine separating sewage systems. Water Science and Technology 38, 17–25. Schouw NL, Danteravanich S, Mosbaek H, Tjell JC (2002) Composition of human excreta – a case study from Southern Thailand. The Science of the Total Environment 286, 155-166. Shiskowski D (2009, September) Global nitrogen management – the role of wastewater management and urine separation as mitigation strategies. A paper presented at WCW Conference & Trade Show, Winnipeg, MB. Sridevi G, Srinivasamurthy CA, Bhaskar S, Viswanath S (2009) Evaluation of source separated human urine (ALW) as a source of nutrients for banana cultivation and impact on quality parameter. Journal of Agricultural and Biological Science, 4(5), 44-48. Lienert J, Larsen TA (2006) Considering user attitude in early development of environmentally friendly technology: a case study of NoMix toilets. Environmental Science and Technology, 40(16), 4838–4844. Wang C, Long R, Wang Q, Liu W, Jing Z, Zhang L (2010) Fertilization and litter effects on the functional group biomass, species diversity of plants, microbial biomass, and enzyme activity of two alpine meadow communities. Plant Soil, 331, 377-389.

Weissenbacher N, Mayr E, Niederberger T, Aschauer C, Lebersorger S, et al. (2008) Alpine infrastructure in Central Europe: integral evaluation of wastewater treatment systems at mountain refuges. Water Science and Technology, 57(12), 20172022.