|
The Use and Efficiency of Constructed Wetlands
in the Riviera Maya
INTRODUCTION
Constructed or created wetlands are systems that are designed to mimic the natural retention, uptake, and decomposition processes of marshlands and mangroves. In the delicate and endangered coastal region of the Yucatan peninsula, where municipal wastewater infrastructure is lacking, constructed wetlands have a sensible and emerging role. This document serves to review the existing function of constructed wetlands, and by assessing their limitations, develop the potential for this technique.
In the Riviera Maya, constructed wetlands are primarily used as an advanced treatment step after a sedimentation or septic tank, for domicile, restaurant, and hotel systems. In several cases they provide tertiary polishing after conventional treatment plants. Since 1996, when the concept and prototype design was introduced to the region by the Planetary Coral Reef Foundation, other companies, professionals, and homeowners have continued to construct these popular ‘ecological’ systems. However, often times they are built without proper guidance. Many wetlands have been negatively affected by unaccounted load, lack of maintenance, and age.
Semarnat Gate Si'an Kaan Bioreserve
Casa del Mar Akumal
A lack of proper understanding has created skepticism among the very government agencies that are needed to support further investigation and regulation of this technique.
METHODOLOGY
Background information was compiled from previous regional research, including that of Dr. Mark Nelson of PCRF, Dr. Edgar Cabrera of the State of Quitana Roo, and Dr. Nancy Hayden’s group from the Univ. of Vermont. Local architects, builders, owners, and groundspeople were also consulted for their experiences.
From April to September of 2004 a visual survey was conducted of 50 constructed wetlands in the Riviera Maya from Akumal to Sian Kaan Bioreserve. Documented parameters included: general state, plant cover and speciation, media type, shape and approximate area, use and management, and discharge method. Six retrofit projects were also observed, where interior processes like sedimentation and root depth could be measured.
A full chemical analysis was completed for seven constructed wetlands, where triplicate samples of both the influent and effluent were measured. In seven other wetlands the effluent was measured at least once. Distinctive discharge sites like cenotes and tertiary filters were also tested. Parameters included: chemical oxygen demand (COD), ammoniacal nitrogen (NH3-N), and phosphate phosphorous (P04-P). Nitrate (NO3) was measured in wetlands with forced aeration. Salinity and pH were periodically monitored using meters. Samples were taken at the same time of day, immediately transferred to the laboratory at Centro Ecológico Akumal, and processed within 4 hours. All tests were performed using Hach colorimetric procedures (Mths 800,8048,10031,8039).
A bottle experiment was completed to measure the sorptive capacity of the gravel on phosphorous. Triplicate bottles with 250g of gravel taken from an excavated wetland of 6 years, used and washed gravel, and new, washed gravel were amended with 450mL of 0, 6.25, 12.5, 25, 50, and 100mg/L P, shaken daily, and the solution sampled throughout a period of 10 days.
Samples were periodically checked for replicability, and one set was verified by an outside laboratory. Treatment comparisons were made using a two-tailed t-test at a significance value of 0.05.
The monitoring scheme was limited by field conditions and laboratory resources. Though these parameters do not represent total levels of nutrients, they represent the majority of the concentration and directly relate in percentage removal. Additionally, the lack of hydraulic flow measurements inhibits accurate conclusions regarding total removal. Due to the water loss through evapotranspiration, nutrient removal is likely to be 10-20% greater than that observed.
RESULTS and DISCUSSION
General State & Design
Regional constructed wetlands show a varying array of shapes, including rectangles, L’s, semi-circles, and irregular forms. They range in gravel depth from 0.5 to 2 meters (1.9 to 6.6 feet) and have an average surface area of 3 m2 per person. Most of the contractors consulted base their design of wetlands for secondary treatment on available space, and the precedent of 3-3.5m2 per person, which corresponds to an average 5-day residence time with other constrained parameters such as water use and cell depth.
The cells are constructed of reinforced concrete block and primarily sealed with concrete stucco which has shown stable results over time. Impermeable additives and liners have also been tried, with less success than the traditional concrete work. The inlet piping is usually 4” perforated PVC running along the width of the cell. Many have 90° elbow clean-outs, which recently have been replaced by 45° or side-access caps for easier cleaning, though they are seldom used.
The majority of the wetlands do not exhibit a larger stone in the inlet zone and principally ľ” washed limestone gravel is used throughout the cell, though unwashed, poor quality gravel is also common. On the older wetlands, irregularities in the surface gravel have been caused by settling and dissolution.
Outlet control structures are usually moveable standpipes that discharge into the subsurface, mangrove, or cenotes. A minority of systems reuse water for irrigation.
Most are on properties whose owners and caretakers know little about the systems. Consequently, they are cared for only when major problems arise, such as flooding and odors.
Plants
Plant selection is usually also based on precedence, and there has been minimal investigation of optimal species since the original survey done by Dr. Mark Nelson and Botanist Edgar Cabrera in 1997. Table 1 indicates the rank in descending order of the most prevalent plants in the 50 regional wetlands.
Avg. root
Rank
Common name
Scientific Name
depth (cm)
Native
1
Taro, elephant ear
Alocasia macrorhiza, Xanthosoma roseum
16
N
2
Papiro, umbrella plant
Cyperus alternifolius
30
N
3
Helecho, fern
Acrostichum danaefolium
28
Y
4
Platano, banana
Musa paradisiacal
26
N
5
Ixora, jungle flame
Ixora coccinea
20
N
6
Platonillo, canna lilly
Canna sp.
12
N
7
Tulipan, hibiscus
Malvaviscus arboreus
15
Y
Table 1. Characteristics of highest ranked plant species.
As shown, the majority of the species are non-native plants with roots that reach less than half of the depth of the wetland cell. Stratification was particularly evident on the walls of excavated wetlands, where the root zone, and area of greatest microbial activity, is seen at the top 25 cm and a dark, anaerobic layer underneath. The US Environmental Protection Agency has also identified this issue in wetlands that not only limits treatment but causes a preferential flow path in the underlying layer of gravel (2000). These effects are also caused by the irregular and incomplete planting exhibited by the average 66% above ground plant cover in the regional constructed wetlands.
Though there is little evidence in the literature to suggest that plant diversity significantly affects performance, most designers agree that biodiversity is an indicator of system health, creating a myriad of niches both within the wetland for internal processes and externally for habitat and beauty. Experience with the two CEA/PCRF wetlands suggests that biodiversity is difficult to sustain, whether naturally evolved or managed. In Wetland I (Nelson 1997) 70% of the species was reduced from 54 species in 1997 to 16 species in 2004. In Wetland II, 83% of the species was reduced from 63 to 11 species. The majority of lost species were understory plants, most notably Typha dominguensis, cattail, initially the highest ranking species, but driven out by shade or competition. It has been the author’s experience that other traditional wetland plants, like bulrushes and reeds, which are used for their extensive root depth, are also difficult to establish in the presence of hard-stemmed species.
The 50 regional wetlands exhibited an average of 5.4 species per system and a Shannon diversity Index (log10) of 0.55, which is slightly higher than that of local mangrove wetlands.
Nutrient Removal
As shown in Fig.1, the average reduction of nitrogen, phosphorous, and chemical oxygen demand are 32%, 1%, and 52% respectively. The corresponding average effluent concentrations are 28.6, 6.3, and 91.6 mg/L. The average influent and effluent pH are 7.36 and 7.31, and the average salinity of both is 5%.
Figure 1. Influent and effluent concentration averages of 7
constructed wetlands, ± std. error of the mean.
Constructed wetlands generally show high levels of performance during system start-up, when plant and gravel uptake is the greatest, and decrease with time. For this reason, it is important to perform long-term monitoring, particularly to determine the design life of systems in this region. The wetlands in Fig. 1 range from 3-7 years. During the first years of operation on two wetlands, Dr. Nelson found average levels of 85% reduction of BOD, 80% reduction in nitrogen, and 78% reduction of phosphorous (1998).
These results indicate that under the current design and operation parameters, constructed wetlands rapidly decrease in performance in a matter of years and do not achieve adequate levels of treatment for discharge in sensitive areas. The US EPA and many other research groups have established that long-term removal of nitrogen and phosphorous is limited, and therefore wetlands should not solely be used in sites where high treatment is needed (2000). For the compromised Mesoamerican Barrier Reef system, nitrogen input is the critical parameter, and wastewater treatment technologies should be designed with this in mind. Unfortunately, little is known about the degradation kinetics of nitrogen in wetlands, other than it usually requires a much longer detention time than 5 days, and is strongly influenced by vegetation patterns and site conditions (Metcalf and Eddy 1991).
Evidence from the regional wetlands also suggests that bed volume and detention time affect nitrogen removal. PCRF/CEA wetland 1, which has a surface area of 16.9m2 per population equivalent, reduced significantly more nitrogen than 3 other regional wetlands of approximately the same age and plant cover, whose capacities range from 3.5-4.5m2 per person. Conversely, these wetlands did not show similar differences in phosphorous removal. The effects of differences in plant cover with nutrient removal could not be extrapolated from the data.
Results from the bottle experiment suggest that although used gravel still exhibits a capacity to adsorb phosphate at high input levels, very little attraction is emitted at those phosphorous levels seen in influent wastewater, of around 6-10mg/L P. Likewise, differences between new, old, and washed gravel are more apparent at high P concentrations (Fig. 2). Washing gravel does partially recharge sites for phosphorous sorption.
Fig. 2. The mass sorption of phosphorous on day 10.
Improving Systems
Several owners and designers have tried to improve system function through additional plantings, cleaning of the inlet and outlet zones, and/or complete excavation. Though adequate (90-100%) plant cover is recommended, it is doubtful that it will create significant effects, until we find and work to establish the species that have maximum root extension, or design much shallower beds, as in the suggested design of the Tennessee Valley Authority of 0.3m. Cleaning of the inlet zone is a simple and worthwhile task, and may be done by uncovering the header pipe and removing the first 0.5-1 m of gravel and relaying larger stone (5-7cm) for initial filtration. Gravel may be added where there are depressions or ponding occurs. The complete excavation of plants and gravel is an expensive and time-consuming task, with only short-term benefits in performance.
The poor performance in carbon and nitrogen removal may be attributed to lack of oxygen, which is a function of the plants, depth, and size of the system, but also due to the lower percentage of oxygen in waters at high temperatures. Three systems have also been amended with air to increase performance. Designed by North American Wetland Engineers, these systems consist of a small electric or solar-powered compressor in a sealed box, which pumps air down to a perforated hose snaked on the bottom of the wetland. In two home systems that run for 12 hours per day, COD was reduced to less than 20mg/L and nitrogen to 0 mg/L. The effluent of these systems showed high levels of nitrate (>20mg/l N), which indicates that the denitrification process to nitrogen gas within the wetland is incomplete, and must be addressed in the discharge method, as nitrate is particularly mobile and deleterious in the environment.
The third aeration system, PCRF/CEA II has encountered technical difficulties since its installation, and lends to the concern of using mechanical systems with natural treatments, especially in regions without the expertise to maintain them.
Aeration does not seem to significantly affect the removal of phosphorous. The main mechanism for reduction is sorption, and the limestone of this region shows a high initial capacity for retention, though may reach saturation as soon as 4 years. Washing and re-using gravel is an option, and though the released ions may be partially retained in the overlying soil or within the limestone aquifer, partial filtration is inevitable (Cable 2002). Current research suggests that the most sustainable phosphorous removal may be through various stages or mixtures of media that are selected for longer-term retention (Brix et al. 2001). CEA and other groups have also begun testing the viability of recyclable material like glass or PET in place of the costly gravel (Dallas 2004).
Recommendations
As an engineered but natural system, this technique must be considered both a science and an art. Local professionals must return to the pure design equations on which biological degradation and hydraulic flow is based, instead of rules based on precedence. These designs will lead us in the ongoing experimentation on the application of this technique to this particular region. Though wetlands require little maintenance, their operation must be maintained, and checked periodically.
The most sensible and achievable role of constructed wetlands in the Riviera Maya is to be part of an integrated system, by providing high-quality water for reuse in surrounding gardens or green space. This is possible on the home, hotel, or municipal level, and would not only avoid contamination, but would use our waste as the resource it is.
References
Cable, Jaye E. 2002. Phosphate Uptake in Coastal Limestone Aquifers: A fresh look at Wastewater Management. Limnology and Oceanography Bulletin, 11(2), p. 29-32.
Dallas, S.C. and Ho, G. 2004. Subsurface Flow Reedbeds Using Alternative Media for the Treatment of Domestic Greywater in Costa Rica. IWA, 6th Specialist Conference on Small Water and Wastewater Systems.
Environmental Protection Agency, USA. 2002. Constructed Wetlands Treatment of Municipal Wastewaters. EPA/625/R-99/010.
Metcalf and Eddy, Inc. 1991. Wastewater Engineering: Treatment, Disposal, and Reuse. McGraw-Hill.
Nelson, Mark. 1998. Limestone Wetland Mesocosm for Recycling Saline Wastewater in Coastal Yucatan, Mexico. Dissertation, University of Florida.
|
