Hospitals, Hotels and Laundries
5.3. Model Development
5.3.1. Data Analysis and Model Development for Restaurants
5.3.1.2. Forecasting Water Use Model Development
Based on the data analysis two different modelling approaches (i.e. average of 2008-2010 water use and 2010 water use) were considered to forecast the annual water use in 2011 for different Restaurants groups and was outlined in Table 5.4. This forecast annual water use was then disaggregated into quarterly water use using the quarterly disaggregating factors. For most of the water user groups, the quarterly disaggregating factors were obtained from the average percentages over 2005-2010 periods as similar quarterly percentages of total annual water use were observed during this period. Further details on water use modelling for individual groups are discussed below.
a. Group: >20 ML
As outlined in Table 5.4, it was found that the annual water use in this restaurant was highly variable. Therefore, the total water use (as it is the recent year’s water use) in 2010 was considered as the forecast total water use in 2011. The disaggregating factors also obtained using the outlined procedure in Table 5.4 and presented in Table 5.5.
Table 5.5 Disaggregating factors for >20 ML Restaurants group
The forecast annual water use in 2011 was then disaggregated into quarterly water use with the factor outlined in Table 5.5. The comparison of the observed and the modelled water use in 2011 are presented in Table 5.6 in both quarterly and annual time steps. The resulted E value from the comparison of the quarterly observed and forecast water use is also presented in the table. The negative E value indicates that the average values would be better predictions than the forecast quarterly water uses.
Restaurant Group
Disaggregating factors for quarters
Quarter 1 Quarter 2 Quarter 3 Quarter 4
Table 5.6 Comparison of observed and forecast water use of >20 ML Restaurants group in 2011
It can be also seen from Figure 5.2, the forecast water use in 2011 was very different from the past annual water use for this restaurant. As this is very high water using restaurant and there is no particular trend can be identified without more information, a future field survey should be considered for predicting the quarterly water use of this restaurant.
b. Group: >15-20 ML
From the data analysis, it can be seen that the total water use in 2008, 2009 and 2010 for this restaurant is close to 15 ML. As the annual water use is almost same in these three consecutive years, the average annual water use of 2008-2010 was consider as the forecast annual water use in 2011 as outlined in Table 5.4. In Table 5.4, it is also outlined that the quarterly percentages of 2005-2010 are similar except those in the year 2009. Therefore, the average quarterly percentages in 2005-2008 and 2010 were considered as the quarterly water use disaggregating factors for this restaurant which is shown in Table 5.7. Finally, the quarterly water use by this restaurant in 2011 was then forecasted using these disaggregating factors and the forecast annual water use in 2011. The observed and forecast water use in 2011 for this restaurant in both quarterly and annual time steps are presented in Table 5.8. The E value from the comparison of the forecast quarterly water uses with the observed quarterly water use in 2011 also presented in the table.
Restaurant Group Annual water use in 2011 (KL)
Water use in each quarter in 2011 (KL) E value 1 2 3 4 >20 ML Observed 25,617 6,871 6,296 6,117 6,333 -0.23 Forecasted 28,375 6,952 6,810 6,952 7,661
Table 5.7 Quarterly disaggregating factors in >15-20 ML Restaurants group
Table 5.8 Comparison of observed and forecast water use of >15-20 ML Restaurants group in 2011
c. Group: >10-15 ML
There are four restaurants in this group. Water use in Restaurant 1 has significantly changed in 2007. Therefore, water use modelling for this restaurant was not performed in this research. Future field survey is recommended to model the water use of this restaurant. As mentioned in Table 5.4, the total water use in 2010 was considered as the forecast annual water use in 2011 for Restaurants 2 and 4. In case of Restaurant 3, from data analysis it was found that the water use was fairly constant over the years. Therefore, the average annual water use in 2008-2010 was considered as the forecast annual water use in 2011 for Restaurant 3.
From data analysis, all these three restaurants were found to have similar quarterly water use percentages for different years except in 2007 in Restaurant 4. Therefore, the average quarterly percentages in 2005-2010 (excluding 2007 in Restaurant 4) were considered as the quarterly water use disaggregating factors individually for all three restaurants as shown in Table 5.9. Thereafter, using the
Restaurant Group
Disaggregating factors for quarters
Quarter 1 Quarter 2 Quarter 3 Quarter 4
>15-20 ML 28 24 22 26 Restaurants Group Annual water use in 2011 (KL)
Water use in each quarter in 2011 (KL) E value 1 2 3 4 >15-20 ML Observed 16,159 4,296 3,500 3,859 4,505 0.22 Forecasted 14,738 4,127 3,537 3,242 3,832
forecast annual water use and these disaggregating factors, quarterly water use in 2011 were forecasted for Restaurants 2, 3 and 4. The E values were also obtained from the comparison of observed and forecast quarterly water use in 2011 for all three restaurants. The E values along with the observed and forecast water use in 2011 for annual and quarterly time steps are presented in Table 5.10. From the table it can be seen that models performed well for Restaurants 2 and 4. However, in case of Restaurants 3, negative E value indicates that the average of previous quarterly water use would be better prediction than the forecast water use in this research.
Table 5.9 Quarterly disaggregating factors in >10-15 ML Restaurants group
Table 5.10 Comparison of observed and forecast water use of >10-15 ML Restaurants group in 2011
>10-15 ML Restaurants
Group
Quarter 1 Quarter 2 Quarter 3 Quarter 4
Restaurant 2 26 25 24 25 Restaurant 3 24 25 26 25 Restaurant 4 26.5 24.5 24 25 >10-15 ML Restaurants Group Annual water use in 2011 (KL)
Water use in each quarter in 2011 (KL) E value 1 2 3 4 Restaurant 2 Observed 9,337 2,583 2,289 2,157 2,309 0.97 Forecasted 9,643 2,507 2,411 2,314 2,411 Restaurant 3 Observed 10,299 2,773 2,739 2,462 2,325 -0.58 Forecast 10,472 2,513 2,618 2,723 2,618 Restaurant 4 Observed 13,592 3,602 3,488 3,428 3,074 0.52 Forecast 14,874 3,942 3,644 3,570 3,718
d. Group: 5-10 ML
There is more than one restaurant in each subgroup in this group (Section 5.3.1.1) and the annual water uses among the restaurants are similar over the years. Therefore, average water use for each subgroup was considered for modelling and forecasting. From data analysis, it was also found that the average water use by the subgroups are similar from 2008-2010. Therefore, average annual water use by each subgroups from 2008 – 2010 are considered as the forecast average annual water use in 2011 for all subgroups.
The quarterly water use disaggregating factors were obtained for each subgroup from the average quarterly percentages excluding for the years mentioned in Table 5.4. These factors are shown in Table 5.11. Finally, the forecast average annual water uses were disaggregated into quarterly water use in each subgroup. The observed and the forecast average quarterly water uses in 2011 for each subgroup are presented in Table 5.12. The E values obtained from the comparison of the forecast quarterly average water uses with the observed quarterly average water uses in 2011 among the subgroups are also presented in Table 5.12. It can be seen from this table that high E
values were obtained among all the subgroups.
Table 5.11 Quarterly disaggregating factors in 5-10 ML Restaurants group
5-10 ML