Verification of Calculation Method Using Monte Carlo Method for Water Supply Demands of Office Building

In Japan, there are four methods of calculating water supply demands for office buildings based on SHASE-S 206 and two methods based on the design standard of Ministry of Land, Infrastructure, Transport and Tourism (MLIT). However, these methods were found to produce overestimated values when applied to recent sanitary fixtures with advanced water saving features. To cope with this problem, Murakawa’s Simulation for Water Consumption (MSWC), which utilizes the Monte Carlo method to calculate water usage dynamically has been developed. In this study, we evaluated the validity of MSWC on water consumption of an office building. Actual water consumption data were collected from a six story office building. Water consumption estimates calculated by the six conventional methods and MSWC were compared with the actual measurement values. Though the calculations based on the conventional methods significantly deviated from the actual measurement values, those made by MSWC closely resembled them.


Introduction
In Japan, the design standard of Ministry of Land, Infrastructure, Transport and Tourism (MLIT) [1] (referred to as "the Design Standard") and The Society of Heating, Air-Conditioning and Sanitary Engineers of Japan Standard 206 [2] (referred to as "SHASE-S 206") have been used since the 1970s as water load calculation methods.However, the traditional water load calculation methods such as the Design Standard and SHASE-S 206, especially the worldwide used Hunter-based method, was proved with a risk of overestimation if they are applied to modern sanitary fixtures with advanced water-saving features, mentioned by Murakawa (1985) [3] and Wu (2013) [4].
On the other hand, Murakawa et al. (1976) mentioned a method to calculate the water supply load based on Monte-Carlo simulation [5,6], and with the known probability of demand and demand flow rate in every hour, daily demand time series can be obtained by Monte-Carlo simulations mentioned by Holmberg (1987) [7], the Monte-Carlo simulations for the calculation of water supply load was considered provide more accurate results.According to that, more research was taken on the Monte-Carlo simulations for water supply loads.Murakawa et al. (2005) [8] developed the simulation tool Murakawa's Simulation for Water Consumption (MSWC), which is based on the Monte-Carlo simulation, enabled to dynamically calculate various water usages in buildings by applying them to method, and Figure 2 shows the calculation flow by the fixture-method.Tables 2 and 3 show each parameter of the personnel method and the fixture-method.
Water 2017, 9, 376 3 of 21  Calculation of daily water consumption by type of personnel: qd (L/d) qd = N•q where N: No. of users by personnel type (persons); q: average daily water consumption per user responding to personnel type (L/(d•persons)).
Calculation of average probable hourly demand of water supply by type of personnel: qh (L/h) qh = qd/t where qd: daily water consumption by personnel type (L/d); t: Daily operating time corresponding to personnel type.
Calculation of probable peak flow rate of water supply: Qm (L/min) where K2: coefficient for peak water consumption (= 1.5-2, usually 2 is used.);Qhm: probable maximum hourly demand of water supply (L/h).
Calculation of Probable maximum hourly demand of water supply by type of fixture: qhm (L/h) qhm = q•n•N where q: water consumption per use by fixture type (L/(frequency/unit)); n: Maximum value of usage frequency per hour corresponding to fixture type (frequency/h); N: Number of fixtures by fixture type (Unit).
Calculation of average probable hourly demand of water supply: Qh (L/h) Qh = Qhm/K1 where Qhm (L/h): Probable maximum hourly demand of water supply (L/h); K1: Coefficient for maximum hourly water consumption (= 1.5-2, usually 2 is used.)Calculation of probable peak flow rate of water supply: Qm (L/min) where K2: coefficient for peak water consumption (= 2-3, usually 2.5 is used.);Qhm: probable maximum hourly demand of water supply (L/h).Calculation of daily water consumption by type of personnel: qd (L/d) qd = N•q where N: No. of users by personnel type (persons); q: average daily water consumption per user responding to personnel type (L/(d•persons)).
Calculation of average probable hourly demand of water supply by type of personnel: qh (L/h) qh = qd/t where qd: daily water consumption by personnel type (L/d); t: Daily operating time corresponding to personnel type.
Calculation of probable peak flow rate of water supply: Qm (L/min) where K2: coefficient for peak water consumption (= 1.5-2, usually 2 is used.);Qhm: probable maximum hourly demand of water supply (L/h).
Calculation of Probable maximum hourly demand of water supply by type of fixture: qhm (L/h) qhm = q•n•N where q: water consumption per use by fixture type (L/(frequency/unit)); n: Maximum value of usage frequency per hour corresponding to fixture type (frequency/h); N: Number of fixtures by fixture type (Unit).
Aggregation of probale maximum hourly demand of water supply: Qhm (L/h) Qhm = qhm1 + qhm2 + qhm3 + … where qhm1, qhm2, qhm3…: probale maximum hourly demand of water supply by fixture type (L/h).Note: * when any hot water faucet is combined, the water supply load unit of fixture for one faucet is calculated as three quarters of the value indicated in this table.

1.
Calculation Method Based on Water Use Time Rate and Fixture Unit for Water Supply Based on the number of the installed fixtures, the Formula (1) and ( 2) by the type of the fixture are used, or Table 4 is used to calculate Q max .The standard value of water use time η and the fixture unit for water supply are shown in Table 5.
In addition, in the case where different types of fixtures coexist, Q max is calculated by adding the highest value among Q max values of each fixture to the half value of Q max of other fixtures.However, continuous flow is to be added without halving its Q max value.
where Y max : maximum number of fixtures used simultaneously with reject rate k (unit); c: number of installed fixtures (unit); ρ: usage rate; η: Water use time rate; b, ∆: constant determined by k.
where Q max : peak flow rate of water supply (simultaneously used water consumption) (L/min); Y max : maximum number of fixtures used simultaneously (unit); q: fixture unit for water supply.

Applicable for various purposes
[Advantages] -Data has been accumulated over many years in this calculation method, which has been, as the result, used for a long time.
-Architects can have many options for their judgement and can calculate pipe diameters based on actual situation.
-Fixed options are established for the case where a number of users is not sufficient for a number of fixtures.
[Disadvantages] -The calculation method requires complex calculation.
-Many provision values are used for calculation of water use time rate η.
-When combining a different type of fixtures, it is required to consider to halve the addition coefficient.
-Adjustment data can be used only for closet bowls, urinals and basins, and adjustment methods for other fixtures are unclear.
(1) Reduction in number of installed fixtures → Maximum number of fixtures used simultaneously Theoretical formula:  -For example, this method is applicable for and effective in the following cases: (1) There are some parts that are used in different time zones in a building, such as a laboratory, room for a cooking class, and students' toilets in a school.
(2) There is a place where time of use is concentrated, such as a shower room in a school.
[Disadvantages] -Applicable only when only a small number of fixtures is used.
( -If a flush valve and flushing water tank coexist, the method calculation is unclear.
-Any consideration is not made to a decrease in the calculation result in the case where the ratio of users to the number of fixtures is low.
-Its safety factor is too high.
(  In addition, if randomly-used fixtures are installed horizontally in 2 or more places or vertically on 2 or higher floors, the number of the fixtures is adjusted to correspond to the number of users by using Table 6.Considering that even if the fixtures are occupied simultaneously, there is still unoccupied time for switch of users, the usage rate was specified as 0.9.Under the assumption that a slight decrease in water pressure due to the simultaneous use is acceptable, the reject rate was specified as k = 0.05, and b = 1.6 and ∆ = 0.8 were employed.Figures 3 and 4 and Table 7 shows relationship between the rejection rate k, and b, and ∆.   2. Method Based on Newer Water Supply Demand Unit [18] The newer water supply demand unit of each fixture is chosen in Table 8 below, and the water load is calculated by using the sum of the newer water supply demand unit of each fixture and the load curve shown in Figure 5.   2. Method Based on Newer Water Supply Demand Unit [18] The newer water supply demand unit of each fixture is chosen in Table 8 below, and the water load is calculated by using the sum of the newer water supply demand unit of each fixture and the load curve shown in Figure 5.   2.
Method Based on Newer Water Supply Demand Unit [18] The newer water supply demand unit of each fixture is chosen in Table 8 below, and the water load is calculated by using the sum of the newer water supply demand unit of each fixture and the load curve shown in Figure 5.

Method Based on Fixture Usage
The water load is calculated by choosing the number of the fixtures used simultaneously in Table 9 and the peak flow rate of water supply of each fixture in Table 10, then by multiplying the number of the fixtures with the peak flow rate.If the number of the fixtures is not indicated in Table 9, the intermediate value is obtained by proportional distribution.However, if usage of the fixture is estimated with high accuracy, the usage rate of the fixtures used simultaneously should be specifically estimated without referring to this table.

Method Based on Fixture Usage
The water load is calculated by choosing the number of the fixtures used simultaneously in Table 9 and the peak flow rate of water supply of each fixture in Table 10, then by multiplying the number of the fixtures with the peak flow rate.If the number of the fixtures is not indicated in Table 9, the intermediate value is obtained by proportional distribution.However, if usage of the fixture is estimated with high accuracy, the usage rate of the fixtures used simultaneously should be specifically estimated without referring to this table.The water load is calculated by obtaining the water supply load unit of each fixture in Table 11 and by using the sum of the obtained water supply load unit of each fixture and the load curve shown in Figure 6.The water load is calculated by obtaining the water supply load unit of each fixture in Table 11 and by using the sum of the obtained water supply load unit of each fixture and the load curve shown in Figure 6.

Outline of MSWC
MSWC is a simulation tool that makes it possible to forecast water supply demand in chronological order by using the Monte Carlo technique.As shown in Figure 7, this method is used for plumbing fixtures, but the frequency ratio of the mean values of the simulation conditions such as use frequency [24], discharge flow rate [25], discharge time of the fixture [26,27], temperature of water to be used [8,[28][29][30], and users' occupation time of the fixture [1,31]

Outline of MSWC
MSWC is a simulation tool that makes it possible to forecast water supply demand in chronological order by using the Monte Carlo technique.As shown in Figure 7, this method is used for plumbing fixtures, but the frequency ratio of the mean values of the simulation conditions such as use frequency [24], discharge flow rate [25], discharge time of the fixture [26,27], temperature of water to be used [8,[28][29][30], and users' occupation time of the fixture [1,31] is cumulated in a probability distribution with a mean value at 1, and can be shown by Erlang distribution, exponential distribution and hyperexponential distribution.In this method, water load can be calculated on a peak-time, hourly and daily basis by generating random numbers and determining the value of each condition with Monte Carlo technique, then by inputting a number of fixtures, average water discharge time, average water discharge quantity and subject number in the subject building used for the calculation.Figure 8 shows the procedure for water load calculation proposed by Murakawa.
Water 2017, 9, 376 11 of 21 probability distribution with a mean value at 1, and can be shown by Erlang distribution, exponential distribution and hyperexponential distribution.In this method, water load can be calculated on a peak-time, hourly and daily basis by generating random numbers and determining the value of each condition with Monte Carlo technique, then by inputting a number of fixtures, average water discharge time, average water discharge quantity and subject number in the subject building used for the calculation.Figure 8 shows the procedure for water load calculation proposed by Murakawa.probability distribution with a mean value at 1, and can be shown by Erlang distribution, exponential distribution and hyperexponential distribution.In this method, water load can be calculated on a peak-time, hourly and daily basis by generating random numbers and determining the value of each condition with Monte Carlo technique, then by inputting a number of fixtures, average water discharge time, average water discharge quantity and subject number in the subject building used for the calculation.Figure 8 shows the procedure for water load calculation proposed by Murakawa.

Calculation Conditions of MSWC
Table 12 shows the simulation conditions of MSWC [4,32].The subject number means a number respectively of male or female users.When calculating the water load of all the floors of the subject building, the number of users in the whole building is to be entered, and when calculating the water load of one floor of the subject building, the number of users on the floor is to be entered.Figure 9 shows the operation screen of MSWC.The indication of each numbered zone on Figure 10 is as follows.In the facility selection section in area " 1 ", the subject building is selected.In this research, "Office" was selected for research of an office building (the subject building can also be selected from the file list on the bottom of the screen by selecting the file and clicking "Read selected file" in area "  12 shows the simulation conditions of MSWC [4,32].The subject number means a number respectively of male or female users.When calculating the water load of all the floors of the subject building, the number of users in the whole building is to be entered, and when calculating the water load of one floor of the subject building, the number of users on the floor is to be entered.Figure 9 shows the operation screen of MSWC.The indication of each numbered zone on Figure 10 is as follows. In the facility selection section in area "①", the subject building is selected.In this research, "Office" was selected for research of an office building (the subject building can also be selected from the file list on the bottom of the screen by selecting the file and clicking "Read selected file" in area "⑪").").
Next, the condition file to be edited is selected to rewrite and store the file.The overall conditions are described as follows.
In the "Analysis unit" section in area " 2 ", it is recommended to choose the 0.1-s analysis, because the 1-s analysis may cause error (taking too long time for the analysis).
In area " 3 ", the result name can be given to the file of the calculation result.If the number exceeding the number of the fixtures chosen in the area " 6 " on the right is entered, error will be caused during the calculation.Hence, the "Fixture type" section should not be changed.The "Subject number" should be consistent with the present condition and be set by considering the situation of the office.
The same operation with the "Set number of fixtures" in area " 4 " can also be performed in the "Fixture name" section in area " 6 " to change the number of the subject fixtures, as well as the "Edit setup number" section in area " 9 ".In the "Set subject number" section in area " 4 ", the number of users and usage of each fixture in 24 h can changed.In the "Set water temperature" section, the temperature of supplied water as well as the water temperature at each fixture can be changed.
In the "Start calculation" section in area " 5 ", the water load calculation simulation based on any calculation condition can be performed by choosing "Rewrite", "Store new file" and the file for the calculation before clicking the "Start".By clicking "Finish" in area " 5 ", data entry can be finished without performing the water load calculation simulation.
In the "Remark" section in area " 7 ", DEMO on the data can be entered.In the "Set value" section in area " 8 ", 11 calculation conditions for each fixture can be shown.In the "Edit set value" section in area " 9 ", 11 calculation conditions for each fixture can be changed.
In the "Progress bar" in area " Water 2017, 9, 376 13 of 21 Next, the condition file to be edited is selected to rewrite and store the file.The overall conditions are described as follows.
In the "Analysis unit" section in area "②", it is recommended to choose the 0.1-s analysis, because the 1-s analysis may cause error (taking too long time for the analysis).
In area "③", the result name can be given to the file of the calculation result.If the number exceeding the number of the fixtures chosen in the area "⑥" on the right is entered, error will be caused during the calculation.Hence, the "Fixture type" section should not be changed.The "Subject number" should be consistent with the present condition and be set by considering the situation of the office.
The same operation with the "Set number of fixtures" in area "④" can also be performed in the "Fixture name" section in area "⑥" to change the number of the subject fixtures, as well as the "Edit setup number" section in area "⑨".In the "Set subject number" section in area "④", the number of users and usage of each fixture in 24 h can be changed.In the "Set water temperature" section, the temperature of supplied water as well as the water temperature at each fixture can be changed.
In the "Start calculation" section in area "⑤", the water load calculation simulation based on any calculation condition can be performed by choosing "Rewrite", "Store new file" and the file for the calculation before clicking the "Start".By clicking "Finish" in area "⑤", data entry can be finished without performing the water load calculation simulation.
In the "Remark" section in area "⑦", DEMO on the data can be entered.In the "Set value" section in area "⑧", 11 calculation conditions for each fixture can be shown.In the "Edit set value" section in area "⑨", 11 calculation conditions for each fixture can be changed.In the "Progress bar" in area "⑩", progress of the calculation can be viewed.By using the "Read chosen file" section in area "⑪", the name of the set value can be shown in the "Set value name" section by choosing and clicking a file in the file list.By using the "Store chosen file" section, the chosen file can be stored.In the "Store new file" section, a new file can be stored in the name used in the "Set value name" section, and the file name can also be changed in this section.
In the "File list" section in area "⑫", the calculation can be performed by checking the box on the left of each file listed as the files that can be used for the calculation.
The water load calculation simulation is conducted for 24 h for each fixture type and is aggregated every second to produce basic files for water, hot water and heat quantity.The simulation is completed, when some of such basic files are displayed.each numbered zone on Figure 10 is as follows.lection section in area "①", the subject building is selected.In this research, or research of an office building (the subject building can also be selected from the f the screen by selecting the file and clicking "Read selected file" in area "⑪").", the name of the set value can be shown in the "Set value name" section by choosing and clicking a file in the file list.By using the "Store chosen file" section, the chosen file can be stored.In the "Store new file" section, a new file can be stored in the name used in the "Set value name" section, and the file name can also be changed in this section.
In the "File list" section in area " Water 2017, 9, 376 13 of 21 Next, the condition file to be edited is selected to rewrite and store the file.The overall conditions are described as follows.
In the "Analysis unit" section in area "②", it is recommended to choose the 0.1-s analysis, because the 1-s analysis may cause error (taking too long time for the analysis).
In area "③", the result name can be given to the file of the calculation result.If the number exceeding the number of the fixtures chosen in the area "⑥" on the right is entered, error will be caused during the calculation.Hence, the "Fixture type" section should not be changed.The "Subject number" should be consistent with the present condition and be set by considering the situation of the office.
The same operation with the "Set number of fixtures" in area "④" can also be performed in the "Fixture name" section in area "⑥" to change the number of the subject fixtures, as well as the "Edit setup number" section in area "⑨".In the "Set subject number" section in area "④", the number of users and usage of each fixture in 24 h can be changed.In the "Set water temperature" section, the temperature of supplied water as well as the water temperature at each fixture can be changed.
In the "Start calculation" section in area "⑤", the water load calculation simulation based on any calculation condition can be performed by choosing "Rewrite", "Store new file" and the file for the calculation before clicking the "Start".By clicking "Finish" in area "⑤", data entry can be finished without performing the water load calculation simulation.
In the "Remark" section in area "⑦", DEMO on the data can be entered.In the "Set value" section in area "⑧", 11 calculation conditions for each fixture can be shown.In the "Edit set value" section in area "⑨", 11 calculation conditions for each fixture can be changed.In the "Progress bar" in area "⑩", progress of the calculation can be viewed.By using the "Read chosen file" section in area "⑪", the name of the set value can be shown in the "Set value name" section by choosing and clicking a file in the file list.By using the "Store chosen file" section, the chosen file can be stored.In the "Store new file" section, a new file can be stored in the name used in the "Set value name" section, and the file name can also be changed in this section.
In the "File list" section in area "⑫", the calculation can be performed by checking the box on the left of each file listed as the files that can be used for the calculation.
The water load calculation simulation is conducted for 24 h for each fixture type and is aggregated every second to produce basic files for water, hot water and heat quantity.The simulation is completed, when some of such basic files are displayed.", the calculation can be performed by checking the box on the left of each file listed as the files that can be used for the calculation.
The water load calculation simulation is conducted for 24 h for each fixture type and is aggregated every second to produce basic files for water, hot water and heat quantity.The simulation is completed, when some of such basic files are displayed.

Measurement of the Water Consumption
Water consumption was measured in an office building to collect basic data, and examine the validity of MSWC and the conventional water load calculation methods.
Data were collected in a 7-story (6 floors above ground and 1 below) office building (referred to as T-building) with the total floor area of 2384.4 m 2 in Tokyo.Figure 11 shows the water supply lines, the locations of ultrasonic flow meters, and the type of the tenants on each floor.Water was supplied by the increase-pressure water supply system.

Measurement of the Water Consumption
Water consumption was measured in an office building to collect basic data, and examine the validity of MSWC and the conventional water load calculation methods.
Data were collected in a 7-story (6 floors above ground and 1 below) office building (referred to as T-building) with the total floor area of 2384.4 m 2 in Tokyo.Figure 11 shows the water supply lines, the locations of ultrasonic flow meters, and the type of the tenants on each floor.Water was supplied by the increase-pressure water supply system.Ultrasonic flow meters were placed in the water supply main near the outlet of the pump (A); in the water supply main between the floors 3 and 4 (B); 4 and 5 (C); and 5 and 6 (D) from Wednesday, 5 August to Thursday, 6 August, and on Thursday, 12 November 2015, and water flow rate was measured every second.Qday and Qmax on each floor obtained from calculation based on the measured data by ultrasonic flow meters during the measuring period: 5-6 August, and 12 November 2015 are shown in Table 13.In Chapter 5, data obtained on 6 August, when Qday calculated from the ultrasonic flow meter data were the largest were used.

The Number of Occupants
To determine the number of occupants present in the rooms in T-building during the measurement period, the occupants of T-building were asked to fill out questionnaires and their Q day and Q max on each floor obtained from calculation based on the measured data by ultrasonic flow meters during the measuring period: 5-6 August, and 12 November 2015 are shown in Table 13.In Chapter 5, data obtained on 6 August, when Q day calculated from the ultrasonic flow meter data were the largest were used.

The Number of Occupants
To determine the number of occupants present in the rooms in T-building during the measurement period, the occupants of T-building were asked to fill out questionnaires and their presence in the rooms every 30 min on the days' measurements were made were investigated.To further validate the accuracy of the questionnaires, security cameras were used [15,16].
The number of occupants in the rooms by gender was counted every 30 min during the measurement period: Wednesday, 5 August to Friday, 7 August; Wednesday, 11 November to Friday, 13 November 2015.
People entering and exiting T-building were monitored by security cameras placed at the front and back entrances during the measurement period: Wednesday, 11 November to Friday, 13 November 2015; and the number of people was counted by gender every 5 min.
The number of occupants registered for T-building is shown in Table 14.The movements of people obtained by questionnaire on Thursday, 12 November and an example of the fluctuation of occupants captured by a security camera on Thursday, November in Figure 12. presence in the rooms every 30 min on the days' measurements were made were investigated.To further validate the accuracy of the questionnaires, security cameras were used [15,16].
The number of occupants in the rooms by gender was counted every 30 min during the measurement period: Wednesday, 5 August to Friday, 7 August; Wednesday, 11 November to Friday, 13 November 2015.
People entering and exiting T-building were monitored by security cameras placed at the front and back entrances during the measurement period: Wednesday, 11 November to Friday, 13 November 2015; and the number of people was counted by gender every 5 min.
The number of occupants registered for T-building is shown in Table 14.The movements of people obtained by questionnaire on Thursday, 12 November and an example of the fluctuation of occupants captured by a security camera on Thursday, 12 November in Figure 12.The maximum numbers of occupants by gender calculated from the data obtained from questionnaires and security cameras are shown in Table 15.Though the fluctuation of occupants was seen both in questionnaire and on security cameras, the female maximum number of occupants monitored on security cameras was about 1.3 times greater than that counted by questionnaire.This may be due to the fact that there were visitors who had not been reflected in the questionnaire.The maximum numbers of occupants by gender calculated from the data obtained from questionnaires and security cameras are shown in Table 15.Though the fluctuation of occupants was seen both in questionnaire and on security cameras, the female maximum number of occupants monitored on security cameras was about 1.3 times greater than that counted by questionnaire.This may be due to the fact that there were visitors who had not been reflected in the questionnaire.

Water Consumption Measurement
Average flush time and average flush volume are two of the simulation conditions set forth in MSWC.Therefore, WC flush time and flush volume were measured to calculate average flush time and average volume in WC in T-building.
Single WC was flushed several times and the fluctuation of flow rate was measured with an ultrasonic flow meter.
Average flush time and average flush volume calculated from five measurements are shown in Table 16.Averages were used for MSWC calculations.Thus, average flush time was 17.2 s and average flush volume 49.8 L/min.(Table 16).17.Actual basic unit was smaller than conventional basic unit in both the personnel method and fixture method.There was a difference of 21,280 L/day in Q day for all floors in the personnel method.

Water Load Calculation Based on MSWC
The simulation conditions in T-building are shown in Table 18 [18].The number of sanitary fixtures in the building, average flush time, average flush volume, and simulation conditions such as the target number were entered to calculate water load.Also, presence rate was calculated from the number of occupants and registrants in the questionnaire, and used as a simulation condition.
The number of registrants on each floor of T-building in August was entered as the target number, and simulations were performed for all floors, B1 to 3rd floor, 4th floor, 5th floor, and 6th floor.The actual number of female multiplied by 1.3 was used as the number of female occupants in this simulation.
The results of simulations in MSWC are shown in Table 19.Except some cases, the simulation results based on personnel/area were larger than those based on the number of occupants, which confirmed that there was a difference of 8713 L/day for all floors.
Average flush time and average flush volume calculated from five measurements are shown in Table 16.Averages were used for MSWC calculations.Thus, average flush time was 17.2 s and average flush volume 49.8 L/min (Table 16).

Comparison of Each Water Load Calculation Method
The Comparison of Q day obtained by each method is shown in Figure 13, the ratio of Q day to actual measurements in Figure 14, the comparison of Q max obtained by each method in Figure 15, and the ratio of Q max to actual measurements in Figure 16 (actual measurements of Q day is referred to as QA day , the ratio of Q day to QA day as Rd, actual measurements of Q max as QA max , and the ratio of Q max to QA max as Rm below).In comparison of Q day , the total figure of each floor and the figures for all floors were used in MSWC.MSWC calculations for the number of occupants were the closest to actual measurements.Compared to the conventional design standards, the personnel method based on the number of occupants and the fixture method based on actual basic unit produced figures closer to actual measurements in Q day and Q max for all floors.However, they were smaller than actual measurements in Q max for floor, indicating that the personnel method and fixture method in the conventional design standards are not reliable when making calculations based on only a few sanitary fixtures installed.
Water 2017, 9, 376 18 of 21 actual measurements in Qday and Qmax for all floors.However, they were smaller than actual measurements in Qmax for each floor, indicating that the personnel method and fixture method in the conventional design standards are not reliable when making calculations based on only a few sanitary fixtures installed.Water 2017, 9, 376 18 of 21 actual measurements in Qday and Qmax for all floors.However, they were smaller than actual measurements in Qmax for each floor, indicating that the personnel method and fixture method in the conventional design standards are not reliable when making calculations based on only a few sanitary fixtures installed.actual measurements in Qday and Qmax for all floors.However, they were smaller than actual measurements in Qmax for each floor, indicating that the personnel method and fixture method in the conventional design standards are not reliable when making calculations based on only a few sanitary fixtures installed.

Conclusions
In this study, the accuracy of water load calculation based on the conventional method and MSWC simulation were compared and validated.
The conventional water load calculation methods were found to overestimate water load except Qmax for each floor in the personnel method and fixture method.Qmax for all floors and Qday in the personnel and fixture methods using actual basic unit produced figures closer to actual measurements than the conventional methods did.
MSWC calculations with the number of occupants were the closest to actual measurements.It was confirmed that obtaining accurate number of people was important as the simulation based on personnel/area produced larger than actual measurements than the simulation based on the number of occupants.The comparison between the conventional methods with and without the correction of the basic fixture unit based on the measurement results proved that the conventional basic fixture unit should be modified [23].
To further refine and validate the accuracy of MSWC, our next step will be to compare the results of each water load calculation method with actual measurements based on detailed measurement of water consumption and counting of the number of people in buildings for multiple uses.

Conclusions
In this study, the accuracy of water load calculation based on the conventional method and MSWC simulation were compared and validated.
The conventional water load calculation methods were found to overestimate water load except Q max for each floor in the personnel method and fixture method.Q max for all floors and Q day in the personnel and fixture methods using actual basic unit produced figures closer to actual measurements than the conventional methods did.
MSWC calculations with the number of occupants were the closest to actual measurements.It was confirmed that obtaining accurate number of people was important as the simulation based on personnel/area produced larger than actual measurements than the simulation based on the number of occupants.The comparison between the conventional methods with and without the correction of the basic fixture unit based on the measurement results proved that the conventional basic fixture unit should be modified [23].
To further refine and validate the accuracy of MSWC, our next step will be to compare the results of each water load calculation method with actual measurements based on detailed measurement of water consumption and counting of the number of people in buildings for multiple uses.

Figure 1 .
Figure 1.Calculation flow by the personnel method.

Figure 2 .
Figure 2. Calculation flow by the fixture-method.

Figure 2 .
Figure 2. Calculation flow by the fixture-method.

Figure 2 .
Figure 2. Calculation flow by the fixture-method.

Figure 3 .
Figure 3. Calculation of Qmax of the Calculation Method Based on Water Use Time Rate and Fixture Unit for Water Supply.

Figure 4 .
Figure 4. Adjustment of number of fixtures (random use).

Figure 3 . 21 Figure 3 .
Figure 3. Calculation of Q max of the Calculation Method Based on Water Use Time Rate and Fixture Unit for Water Supply.

Figure 4 .
Figure 4. Adjustment of number of fixtures (random use).
Peak flow rate of water supply (L/min) Number of

Figure 4 .
Figure 4. Adjustment of number of fixtures (random use).

Figure 5 .
Figure 5. Calculation of Q max of the Method Based on Newer Water Supply Demand Unit [2,21,22].

Figure 6 .
Figure 6.Calculation of Qmax of the Method Based on Water Supply Load Unit of Fixture [2,23].(a) Peak flow rate of water supply [L/min]; (b) Peak flow rate of water supply [L/min] (Partial expansion of (a)).

Figure 6 .
Figure 6.Calculation of Q max of the Method Based on Water Supply Load Unit of Fixture [2,23].(a) Peak flow rate of water supply [L/min]; (b) Peak flow rate of water supply [L/min] (Partial expansion of (a)).

Figure 7 .
Figure 7. Probability and frequency distribution of each item.

Figure 7 .
Figure 7. Probability and frequency distribution of each item.
Note: As there is no English version of MSWC at present, the right figure was translated depends on the left one.

Figure 10 .
Figure 10.The operation screen of MSWC.Figure 10.The operation screen of MSWC.

Figure 10 .
Figure 10.The operation screen of MSWC.Figure 10.The operation screen of MSWC.

Figure 12 .
Figure 12. Results of the number of occupants.

Figure 13 .
Figure 13.Comparison of Qday obtained by each method.

Figure 14 .
Figure 14.Ratio of Qday to actual measurements.

Figure 13 .
Figure 13.Comparison of Qday obtained by each method.

Figure 14 .
Figure 14.Ratio of Qday to actual measurements.

Figure 13 .
Figure 13.Comparison of Qday obtained by each method.

Figure 14 .
Figure 14.Ratio of Qday to actual measurements.

Figure 16 .
Figure 16.Ratio of Q max to actual measurements.

Table 2 .
Volume and time of water consumption per person based on intended use of building.When the actual number is given, that number is used.However, a future increase is to be projected; *2The office includes a president's office, secretarial office, directors' office, conference room and reception room.Remarks: (1) except as otherwise noted in the remarks section, make-up water in the cooling tower and water consumption in workers' cafeteria are added separately; (2) if caretaker or other maintenance staff is permanently stationed, water consumption by him or her is added separately.For calculation of the water consumption by him or her, values for apartment houses are applied.

Table 3 .
Volume and time of water consumption per person based on intended use of building.

Table 4 .
Classification of calculation method of peak flow rate of water supply.

Table 5 .
Fixture unit for water supply and standard value of water use time η.

Table 6 .
Number of fixtures and usage rate.
Note: (a) where c = 1 in the usage rate approximation ρ = 0.1684ln(c) − 0.0087, the actual calculated figure is −0.01, but is replaced with 0 in this research.

Table 7 .
Number of fixtures and usage rate.

Table 7 .
Number of fixtures and usage rate.

Table 7 .
Number of fixtures and usage rate.

Table 8 .
Newer Water Supply Demand Unit of an Office Building.

Table 8 .
Newer Water Supply Demand Unit of an Office Building.
Figure 5. Calculation of Qmax of the Method Based on Newer Water Supply Demand Unit [2,21,22].

Table 9 .
Number of fixtures and usage rate.

Table 9 .
Number of fixtures and usage rate.

Table 10 .
Usage rate and Q max of each plumbing fixture and faucet.

Table 10 .
Usage rate and Qmax of each plumbing fixture and faucet.
Subject number (No. of users, tenants and rooms) To be entered based on the subject building Water 2017, x, x 7 of 22 Note: As there is no English version of MSWC at present, the right figure was translated depends on the left one.
Subject number (No. of users, tenants and rooms)To be entered based on the subject building

Table 13 .
Qday and Qmax on each floor.

Table 13 .
Q day and Q max on each floor.

Table 14 .
The number of occupants registered for T-building.

Table 14 .
The number of occupants registered for T-building.

Table 15 .
The maximum numbers of occupants.

Table 15 .
The maximum numbers of occupants.

Table 16 .
Average flush time and average flush volume.

Table 17 .
Q day and Q max obtained from calculations.

Table 18 .
Simulation conditions in T-building.

Table 19 .
Results of simulations in MSWC.