Figure 2(a) shows an overview of the lagoon system with screens and curved shapes. Figure 2(b) shows a schematic with 21 sampling sites, where site 1 is located at the entrance in lagoon system and site 21 is located at output in lagoon system.

Lagoon system is built with porous block, as it aims to build such facilities in rural areas for economic reasons and on the other hand it is to take advantage of porosity of the material for fixing bacterial film.

Sampling in each sites selected was monitored for 2 periods: winter and spring 2016, at the same time from 10:00 a.m. to 4:00 p.m. with the purpose of ensure microbial activity. Sampling was carried out at a regular depth of 25 cm with kind bailer samplers at 13:00 hours.

In order to determine the behavior of lagoon system as a function of parameters: temperature, pH, conductivity, redox potential (ORP), dissolved oxygen (DO), total solids suspended (TSS) sites in seven mentioned channels were defined. pH was determined according to the NMX-AA-008-SCFI-2000 [16] standard, while temperature was determined according to NMX-AA-007-SCFI-2000 [17] . For each parameter a Hanna model HI 1618 sensor was used, with a Hanna HI98150 interface. Finally, for conductivity and ORP measurements a LabQuest Vernier Interface system with its respective sensors was used.

Alkalinity [18] and Hardness was determined [19] . The COD was determined by Hach method with a Hanna brand reactor model HI-839800 and a UV/visible spectrophotometer DR/2400 brand Hach. To obtain total solids (TSS) the technique of standard methods [20] was followed. The volume collected in all cases was one litter in Nalgen containers, kept under refrigeration at 4˚C. In situ measurements, for maximum dissolved oxygen concentration, corresponding to a maximum microbiological activity was performed in all samples at 1:00 p.m. [21] . All determinations were performed on an average of two samplings per week per season (winter and spring).

It was important to consider optimization and homogenization of water flow in the input of lagoon system, for which landfills were designed, using Thomson Equation (1) for triangular landfills [22] .

for landfills of 90˚ (1)

where:

Q = Flow (m³/s).

H = Water from the landfill (m).

Flow rate that enters in lagoon system from wetland was Q = 2 L/s, and for design of landfill was considered from Thomson Table [22] a coefficient of 1.4, obtaining a total load H equal to 0.867 cm, which was divided among 12 landfills, so that H/landfill of 0.072 cm was obtained (Figure 3).

For hydrodynamic study in lagoon system, IBER software was used [23] which is a numerical model of simulation of free-flow in non-permanent regime and of environmental processes in fluvial hydraulics. Application of IBER covers fluvial hydrodynamics, flood zones evaluation, transport of suspended material and aquatic flows calculation. Hydrodynamic module for two dimensions was used for this study, where conservation equations of mass and momentum were solved.

where:

h = depth;

U_x, U_y = Horizontal velocities averaged in depth;

G = Acceleration of gravity;

Z_s = Free surface elevation;

τ_s = Friction in free surface due to the friction produced by wind;

ρ = Density of water;

Ω = Angular velocity of rotation of the earth;

Λ = Latitude of the point considered;

, , = Horizontal effective tangential tensions;

M_s, M_x, M_y = Source/sink sum of mass and moment, through which the modelling is performed.

In Figure 4, it is observed a higher concentration of solids (mg/L) in every channel change (3 - 4, 6 - 7, 9 - 10, 12 - 13, 15 - 16 and 18 - 19), due to the entrance of water flow in each curve with slope, phenomenon that occurred in both periods.

Overall, a tendency to decrease in concentration of TSS was observed, being scarce on departure at exit of lagoon system, which indicates a homogeneous sedimentation process along each channel.

Dissolved Oxygen (DO) variation in lagoon system is observed in Figure 5. It is

important to mention that water which entering in system, comes from a treatment sub-surface wetland of horizontal flow [14] later enters in lagoon through 12 landfills, where flow is distributed and conducted to first channel in lagoon system, a jump of 50 cm height allowing aeration of water is produced. DO oscillate in values 3.04 ± 0.144 mg/L at the entrance of first channel during the winter and 3.00 ± 0.132 in spring. DO concentration increases as water flow along each channel is advanced, later it is reduce to input of each curve with slope, and is indicated by vertical dotted lines.

Finally, the concentration improves at the exit of each curve and during straight sections. DO behavior is consistent with the redox potential (ORP) (Figure 6) and COD (Figure 7). Oxygenation was better in winter due to presence of algae and also due to lower evapotranspiration. In spring, oxygenation was lower despite presence of algae and an increase in temperature was observed. Due to action of wind and rain season during April and May respectively, there was a significant fall of leaves, reason for which degradation of organic matter could have consumed part of oxygen. Overall a decrease of DO is observed, it was not so abrupt in first channels, and in spite of this the behavior profiles were retained.

In Figure 6, it was observed that ORP approached increasingly positive values as the flow advanced from channel to channel, behavior that coincided with DO variation. In winter, the ORP presented values more negative than in the spring, but from site 5 (corresponding to channel 2) acquired values slightly more positive than those of spring, behavior that lasted until the exit of the system.

It is also observed that at site 1 (entrance to the system) is the site with the most negative value in both stations, as the flow advances along the channels, an improvement in conductivity is observed, behavior that coincides with the Improvement of DO and COD. It can also be observed that at the beginning of each ORP curve it acquired negative values.

Figure 8 shows results using the two graphs according two stations, a very marked trend of decreasing in BOD was observed. At entrances of curves shapes with

slope, BOD increased, due to the fact that there was more accumulation of biodegradable material, the degradation was increasing until its decrease in the out- put of lagoon system.

COD is a representative measure of organic contamination in an effluent, parameter which will be controlled within the different discharge regulations. Figure 7 presents results of COD as a parameter for monitoring the degradation of organic material, both biodegradable and non-biodegradable in the different channels along the lagoon system. There was a general trend of decreasing COD along the channels at both stations. At the beginning of each banked curve a higher concentration of COD was presented, where more oxygen was required for oxidation of organic matter and it is related with ORP values because these parameters are closely related.

In Figure 9(a), the result of modeling with IBER software is presented, observing a velocity distribution in different sections of lagoon system, identifying the variation of water flow with different colors: light blue means a uniform flow and changes to dark blue in external curve section, where the flow decreases its velocity. In internal curve section there is an increase that impels it to exit of the curve, stabilizing this flow along of channel. The phenomenon is repeated in each curve until last channel, where flow acquires a greater velocity until output of lagoon, which is represented with yellow-green color (Figure 9(a)). Note the almost yellow color where flow is virtually free.

Froude number was determinate, which relates forces of inertia and gravity, implicitly considering velocity and hydraulic gradient at each point of path the water flow, evaluating type of regime. Figure 9(b) shows increasing Froude number as the water flow increased in each channels, ending with a tendency toward a critical regime in the seventh channel. It can also be observed, as in velocity case

that increase at the end of each of curves, which is due to hydraulic jump caused by slope change in curves. A change in critical regime was presented, for example, in changes in slope. The flow in critical regime (or in its environs) was unstable [14] .

This software was also used to model TSS transport in the lagoon system (Figure 10). The concentration of TSS in water is corresponding to field measured values which were introduced in the software. The results indicate that the highest concentration occurred in the first channel (red color), although it seems that before the entrances to the curves the TSS tend to concentrate, this can be due to reduction velocity. In the other channels also there is a drag of TSS, always concentrating at the entrance of each curve, until concentration decreases. It is important to note with this model that in zones of the curves, blue color is homogeneous, which indicates that there are no short circuits, therefore there are no “dead zones” and the hydraulic flow improves significantly.

Table 2 summarizes BOD and COD values as parameters for efficiency follow

up in each channel. Regarding BOD in the first and seventh channels, the best efficiency was observed in the sixth and seventh channels for COD.

Finally in Table 3, efficiency of all system is shown, considering Figure 1. In pre-treatment, practically the same efficiency was achieved in two seasons with a little higher value in winter for the three parameters. In the wetland, TSS concentration was slightly higher in winter, however BOD and COD were better in spring, possibly the rise of temperature improved the ability of microorganisms to degrade MO.

In case of lagoon system, the treatment for TSS was better in winter, because fewer solids were introduced in wetland. The BOD was slightly better in winter, where the flow velocity can be influenced by the curved forms of each channel that avoid short circuits and therefore dead zones, same can be said of COD.