The variability of the 1,000 series of daily rainfall was verified and a similarity in the parameter’s values among the parameters from them and the 25-year historical data was found. This indicates that these rainfall may not be of great impact in any differences in the feasibility of the RWHSs which can be explained by the method used to create the series.
The ideal tank capacity defined using technical and economic performance criteria was evaluated, Fig. 4 in Appendix illustrates the number of scenarios in which the RWHSs performed well for different criteria in Rio Verde and the results for other cities are quite similar.
Evaluations of RWHSs using only technical indicators resulted in large sizes of storage tanks impacting the economic analysis. With the efficiency indicators it was possible to visualize that for 1%/m³ the most successful system required storage tanks of 10 and 15 m³ and for 4%/m³ the most successful system required storage tanks of 10 and 15 m³. As concluded in Pacheco and Alves (2021), changing the efficiency value of performance criteria impact the definition of the size of the storage tank and it shows that different values of this indicator must be used for each level of demand.
The maximum BCR and the NPVV criteria suggested sizes of storage tank smaller than 5 m³ for most demands, while the NPV criteria suggested different volumes. Then, the NPV generates intermediate economic and technical indicators and this criteria was applied in this work to select the size of the storage tanks.
The overall performance of the RWHSs was evaluated upon the analysis of the number of scenarios in which the system performed well in terms of the economic and technical criteria defined. Initial results in Table 6 show the number of favorable scenarios among the 16 thousand scenarios in each city for different criteria thresholds. The threshold for the economic criteria was set to be equal or greater than zero for NPV and NPVV and equal or greater than one for the BCR criteria.
The threshold for the technical indicators varies from 30–50% for RH and from 30–90% for REL and SD indicators. The first one does not have great impact in the feasibility decision, it is an indicator that represents the availability of rainfall that is actually harvested. The two last indicators are more representative of the feasibility of the system. In present work, the RWHSs are alternative source of water supply for non-potable uses, a backup system. This purpose leads to accept a level of technical performance usually unacceptable for a water resource system. Additionally, REL indicator counts favorable only when the non-potable use is fully met at that day.
The RWHSs for Rio Verde presented the largest number of favorable scenarios according to all indicators individually, except for RH in which Formosa was the most successful city. The grey boxes in Table 6 indicates the city with greater number of favorable scenarios for different criteria. One can observe that depending on different threshold for the set of criteria, Rio Verde and Formosa may present the largest number of favorable scenarios. While Ipameri ranks an intermediate position between them.
Given the preliminary results in Table 6 and the similarity of the RWHSs configurations, the better performance of Rio Verde might be due to the rainfall regime in this city that shows better climate indicators.
Table 6
Number of favorable scenarios for different performance criteria
Thresholds | Number of favorable scenarios |
Economic criteria | Techinical criteria |
SD (%) | REL (%) | RH (%) | Formosa | Ipameri | Rio Verde |
ok | - | - | - | 10,044 | 10,317 | 10,741 |
- | ≥ 50 | - | - | 7,741 | 9,824 | 11,659 |
- | - | ≥ 50 | | 6,736 | 8,537 | 10,307 |
- | - | - | ≥ 40 | 7,033 | 7,010 | 6,389 |
ok | ≥ 30 | ≥ 30 | ≥ 30 | 9,414 | 9,307 | 9,119 |
ok | ≥ 50 | ≥ 50 | ≥ 30 | 6,082 | 7,390 | 8,001 |
ok | ≥ 70 | ≥ 70 | ≥ 30 | 1,863 | 2,280 | 2,457 |
ok | ≥ 90 | ≥ 90 | ≥ 30 | 379 | 459 | 513 |
ok | ≥ 30 | ≥ 30 | ≥ 40 | 6,306 | 6,274 | 5,673 |
ok | ≥ 50 | ≥ 50 | ≥ 40 | 4,691 | 5,169 | 5,213 |
ok | ≥ 70 | ≥ 70 | ≥ 40 | 1,851 | 2,236 | 2,281 |
ok | ≥ 90 | ≥ 90 | ≥ 40 | 368 | 457 | 510 |
ok | ≥ 30 | ≥ 30 | ≥ 50 | 3,963 | 3,187 | 2,766 |
ok | ≥ 50 | ≥ 50 | ≥ 50 | 2,643 | 2,593 | 2,462 |
ok | ≥ 70 | ≥ 70 | ≥ 50 | 1,215 | 1,196 | 1,063 |
ok | ≥ 90 | ≥ 90 | ≥ 50 | 193 | 163 | 204 |
ok | ≥ 50 | ≥ 30 | - | 7,168 | 8,357 | 9,639 |
ok | ≥ 50 | ≥ 50 | - | 6,299 | 7,969 | 8,972 |
ok | ≥ 60 | ≥ 30 | - | 3,045 | 4,530 | 5,260 |
ok | ≥ 60 | ≥ 60 | - | 2,920 | 4,182 | 4,582 |
ok | ≥ 70 | ≥ 30 | - | 2,060 | 2,516 | 2,693 |
ok | ≥ 70 | ≥ 70 | - | 2,031 | 2,460 | 2,646 |
Figure 5 presents the performance of the system for all SOW considering the economic and technical indicators for a predefined criteria threshold of SD ≥ 50%, REL ≥ 50% and RH ≥ 30%, one can see that the SD indicator is greater or equal to the REL for all scenarios in the three cities. Considering the previous criteria threshold, the number of favorable scenarios are 8,001 (50%), 7,390 (46,19%) and 6,082 (38,01%) in Rio Verde, Ipameri and Formosa, respectively. It is reasonable to consider the threshold of 50% for the SD and REL indicators once these RWHSs operate as an alternative source of water supply.
The performance was also evaluated for each of the eight water demand categories in Formosa and Fig. 6 of Appendix shows the results. It indicates that is more difficult to meet higher technical criteria thresholds for higher level of water demand. In general, for higher water demand the economic indicators are more favorable, once there is great advantage to use water. However, the amount of water actually available is not enough for the total water demand preventing the systems to reach high levels of REL and SD, technical indicators.
Figure 7 shows the number of favorable scenarios according to different performance criteria and RWHS designs. Both SD and REL criteria are considered satisfactory at the level of 40% to RWHS supplying the demands 1, 6, 7 and 8. Analysis for the RH criteria shows great variability of performance among the RHWS configurations. Systems with high levels of demand and largest roof areas performed better, probably due to larger sizes of accumulation tanks defined for these systems. Regarding the economic criteria, greater NPVs corresponds to progressive increase of performance for RWHS systems at greater demand levels and roof areas. The analysis of combined criteria, technical and economic, reduces the number of favorable scenarios for most of the RWHS configurations. When RWHS are considered as an alternative of water supply, low levels of REL and SD criteria are acceptable, specially observing the economic restriction (criteria) to build larger accumulation tanks. These criteria better represent the contribution of RHWS to enhance the water security in urban areas. These results indicate that RWHS configurations with demands 6, 7 and 8 perform better in changing scenarios.
Figure 8 illustrates the influence of the DU factors in selected performance criteria of RWHSs in the city of Rio Verde for four selected categories of water demand. Analysis using data from other cities showed similar results.
According to the graphs, lower levels of discount rates and higher levels of tariffs contribute to higher values of NPV. On the other hand, the variability of operational and maintenance cost seems not to influence the performance of the systems.
Higher values of NPV are associated to lower values of SD that also decreases for higher water demand. Additionally, the sizes of the circles indicate that the bigger storage tanks are more frequently associated to smaller SD indicators.
The impact of some rainfall series characteristics also was evaluated for NPV and SD performance indicators and there is no significant influence of the rainfall parameters on the performance of the RWHSs, once there are favorable results along all of the range of parameters. This might be due to the analysis be related to rainfall series from the same site and resulted from a bootstrapping sampling.