Water consumption

Objective

They aimed to observe and analyze Two major entities, that is, water demand and supply within machakos municipality.

??
From the questionnaire we can only extract infor on consumption/demand

The research aimed to establish the current water demand and supply and subject the two units of analysis into possible current or future scenarios and draw conclusions from the results.

3.3.2 Variable specifications and variable indicators.

“The main variables in this research shall be water demand and supply. The change in water demand shall be indicated by demographic variation within Machakos Municipality while that of the water supply shall be indicated by any variation caused to the water supply system of Machakos Municipality.”

This is how they define the main outcome var. But am not sure how this is captured in the questionnaire?!!. What has been captured is water demand using the following question:-

Average number of persons needing water per residence:-
Type of water supply do you have!!!
Water Consumption in the past 3 months

The only question i thought could give us an insight on the water supply is:-

Number of months that sources dry up.

But this only talks of number of months. so not sure how to quantify that.!!!!

I am looking at a variable probably talking of the amount of water supplied/available at the water sources so we can try model this bis a bis the demand.

My thought..

The only thing we can try do now is see how the different factors collected affect/influence the amount of water demanded!!!. That is, get a list of factors that affect water consumption.

Unfortunately not sure we can answer this:-

Information about the degree by which the water demand within Machakos Municipality has been met using the current existing water resources systems

Some demographics

Data was collected from 301 respondents from 4 different places in Machakos County. Each of the 4 places contributed \(\approx 25\%\) of the respondents. 51.7% of the respondents were females while 46.7% were male of age ranging from \(<18yrs\) to \(>30yrs\) . Table below shows the Age distribution of the respondents

Age Distribution
	Frequency	%(NA+)	%(NA-)
18-30 yrs	165	54.6	56.5
>30 yrs	117	38.7	40.1
<18 yrs	10	3.3	3.4
NA’s	10	3.3	0.0
Total	302	100.0	100.0

Village/Town Distribution
	Frequency	%(NA+)	%(NA-)
Machakos	198	65.6	67.3
Machakos Katoloni	42	13.9	14.3
Miwani	33	10.9	11.2
Kenya Israel	15	5.0	5.1
NA	8	2.6	0.0
Machakos Kathemboni	6	2.0	2.0
Total	302	100.0	100.0

Level of Education
	Frequency	%(NA+)	%(NA-)
Above secondary	132	43.7	44.3
Secondary	106	35.1	35.6
Primary	60	19.9	20.1
NA’s	4	1.3	0.0
None	0	0.0	0.0
Total	302	100.0	100.0

Occupation
	Frequency	%(NA+)	%(NA-)
Employed	143	47.4	48.6
Self-employed	121	40.1	41.2
None	30	9.9	10.2
NA’s	8	2.6	0.0
Total	302	100.0	100.0

Category of Water Use
	Frequency	%(NA+)	%(NA-)
House-hold	126	41.7	42.3
Shop	49	16.2	16.4
Other	32	10.6	10.7
Hotel	31	10.3	10.4
Bar/restaurant	23	7.6	7.7
Butcher	22	7.3	7.4
School	8	2.6	2.7
Hospital	7	2.3	2.3
NA	4	1.3	0.0
Total	302	100.0	100.0

With the ‘Other’ category expanded

Category of Water Use. With ‘Other’ Category expanded
	Frequency	%(NA+)	%(NA-)
House-hold	126	41.7	42.3
Shop	49	16.2	16.4
Hotel	31	10.3	10.4
Bar/restaurant	23	7.6	7.7
Butcher	22	7.3	7.4
Other: Salon	15	5.0	5.0
School	8	2.6	2.7
Hospital	7	2.3	2.3
NA	4	1.3	0.0
Other: Church	3	1.0	1.0
Other: Barber Shop	2	0.7	0.7
Other: Office	2	0.7	0.7
Other: Welding	2	0.7	0.7
Other: Cereals Shop	1	0.3	0.3
Other: Cyber and Salon	1	0.3	0.3
Other: Cybercafe	1	0.3	0.3
Other: Dry Cleaner	1	0.3	0.3
Other: NA	1	0.3	0.3
Other: Organization	1	0.3	0.3
Other: Pharmacy	1	0.3	0.3
Other: Vegetable Vendor	1	0.3	0.3
Total	302	100.0	100.0

Average number of persons needing water per residence.

Average Number of Persons Needing Water/Residence
Min	Median	Mean	Max
1	5	46	1000

Type of water supply do you have.

Water Supply Type
	Frequency	%(NA+)	%(NA-)
Private Borehole	107	35.4	35.7
Other	87	28.8	29.0
Municipal water supply	71	23.5	23.7
Shallow well	17	5.6	5.7
Rain water harvested from the roof	12	4.0	4.0
Dam/Earth pan	5	1.7	1.7
NA	2	0.7	0.0
Spring	1	0.3	0.3
Total	302	100.0	100.0

Summary of the water consumption in the past 3 months kes.

Water Consumtion in the past 3 months
	Units consumed(??)			Amount(kes)
Month	Mean	Median	[Min, Max]	MeanA	MedianA	[MinA, MaxA]
June	5891.811	2450	[60, 70000]	1734.256	600	[0, 60000]
July	5809.076	2500	[60, 70000]	1755.225	600	[0, 60000]
Aug	5845.929	2400	[0, 70000]	1675.684	600	[0, 40000]

Distance from the main water source to the respondents residence.

Distance(km) from Main water source: Wet season
	Frequency	%(NA+)	%(NA-)
0-0.5	205	67.9	68.6
0.5-1	49	16.2	16.4
1-2	25	8.3	8.4
2-5+	20	6.6	6.7
NA	3	1.0	0.0
Total	302	100.0	100.0

Distance(km)

Distance(km) from Main water source: Wet season
	Frequency	%(NA+)	%(NA-)
0-0.5	122	40.4	41.4
0.5-1	119	39.4	40.3
1-2	33	10.9	11.2
2-5	19	6.3	6.4
NA	7	2.3	0.0
>5	2	0.7	0.7
Total	302	100.0	100.0

Time taken to draw water from the main water source to your place.

Time(mins) taken from Main water source: Wet season
	Frequency	%(NA+)	%(NA-)
<15	228	75.5	76.0
15-30	65	21.5	21.7
30-60	7	2.3	2.3
NA	2	0.7	0.0
Total	302	100.0	100.0

Time(Minutes) taken from water source.

Time(mins) taken from Main water source: Dry season
	Frequency	%(NA+)	%(NA-)
<15	144	47.7	49.0
15-30	138	45.7	46.9
30-60	12	4.0	4.1
NA	8	2.6	0.0
Total	302	100.0	100.0

Months in the year that the main water source dry up.

Number of months that sources dry up
	Frequency	%(NA+)	%(NA-)
0	223	73.8	74.6
2	33	10.9	11.0
1	31	10.3	10.4
3	11	3.6	3.7
NA	3	1.0	0.0
4	1	0.3	0.3
Total	302	100.0	100.0

The Quality of water at the main source.

Quality of Water
	Frequency	Percent	Cum. percent
CLEAR	115	38.1	38.1
SALTY	82	27.2	65.2
SALTY,CLEAR	73	24.2	89.4
SALTY,BROWN	9	3.0	92.4
BROWN	9	3.0	95.4
SALTY,BROWN,CLEAR	3	1.0	96.4
BROWN,CLEAR	3	1.0	97.4
	3	1.0	98.3
SALTY,SMELLY,CLEAR	1	0.3	98.7
SALTY,BROWN,SMELLY	1	0.3	99.0
GREEN,CLEAR	1	0.3	99.3
GREEN	1	0.3	99.7
BROWN,GREEN	1	0.3	100.0
Total	302	100.0	100.0

Results.

Objective 1.

To identify factors that affect water consumption patterns in Machakos Municipality.

To achive this objective a regression model was used to try and find out the factors that affect/influence water consumption. The approach used to look at Water demand in the municipality defined consumption by the number of units consumed in the past 3 months.

The response/outcome variable was defined as follows:-

A total sum of the consumption units for the 3 months.

The table below displays a summary of the model fit statistics for the simple regression models/univariate analysis. From the many predictors analysed only those with a pvalue \(<0.25\) will be considered for the multiple regression model. From the summary table average number of person in a residence seem to be the most informative predictor based on the amount of variation in the outcome variable it explains(Adj R-squared).

Summary for the univarites
r.squared	adj.r.squared	sigma	statistic	p.value	df	logLik	AIC	BIC	deviance	df.residual	predictor
0.4901412	0.4884245	20168.70	285.5142161	0.0000000	2	-3386.914	6779.827	6790.928	120812613473	297	avepersonsatresidence
0.2331755	0.2146660	25132.40	12.5975759	0.0000000	8	-3438.099	6894.198	6927.472	183174866574	290	waterusecategory
0.0985412	0.0864004	26958.27	8.1164947	0.0000032	5	-3507.016	7026.032	7048.295	215844219614	297	place
0.0311299	0.0278456	27999.65	9.4783807	0.0022747	2	-3461.686	6929.371	6940.453	231274144434	295	gender
0.0355487	0.0257407	27952.99	3.6244671	0.0134897	4	-3483.495	6976.989	6995.491	230503977271	295	distwetseason
0.0418458	0.0286299	28054.34	3.1663200	0.0143683	5	-3437.426	6886.851	6908.973	228243415002	290	distdryseason
0.0503938	0.0309479	27831.80	2.5914926	0.0183765	7	-3492.320	7000.641	7030.271	226960517935	293	watersupplytypea
0.0166078	0.0133078	28083.98	5.0327014	0.0256076	2	-3497.565	7001.129	7012.240	235035541896	298	metered_water
0.0217603	0.0150370	28298.73	3.2365541	0.0407181	3	-3429.328	6866.657	6881.391	233038024510	291	occupation
0.0536415	0.0143464	28001.17	1.3650921	0.1819386	13	-3514.356	7056.712	7108.658	226594951463	289	quality1
0.0041388	0.0007970	28261.46	1.2385017	0.2666569	2	-3499.455	7004.909	7016.020	238015679048	298	payforwater
0.0082952	0.0014322	28589.11	1.2086852	0.3000944	3	-3408.970	6825.941	6840.648	236210509937	289	age
0.0061405	-0.0005975	28365.61	0.9113204	0.4031217	3	-3476.710	6961.420	6976.208	237359229107	295	education
0.0092746	-0.0042047	28379.51	0.6880627	0.6007126	5	-3487.515	6987.030	7009.232	236786589220	294	monthssourcedry
0.0033448	-0.0035051	28548.28	0.4882981	0.6141718	3	-3431.910	6871.819	6886.554	237166222852	291	timedryseason
0.0022338	-0.0044851	28336.07	0.3324669	0.7174190	3	-3499.741	7007.482	7022.298	238470986117	297	timewetseason
0.0066406	-0.0071084	27990.11	0.4829874	0.7482334	5	-3425.091	6862.181	6884.283	226415966252	289	villag_town

Predictors selected for the multiple regression include place, gender, occupation, waterusecategory, avepersonsatresidence, watersupplytypea, metered_water, quality1, distdryseason, distwetseason.

Start:  AIC=5656.05
unitsTotal ~ place + gender + occupation + waterusecategory + 
    avepersonsatresidence + watersupplytypea + metered_water + 
    quality1 + distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- quality1              11 3.5603e+09 9.9115e+10 5644.4
- watersupplytypea       6 1.6061e+08 9.5715e+10 5644.5
- occupation             2 4.8763e+08 9.6042e+10 5653.5
- gender                 1 2.7783e+07 9.5582e+10 5654.1
- metered_water          1 3.7378e+07 9.5592e+10 5654.2
- distwetseason          3 1.6032e+09 9.7158e+10 5654.8
<none>                                9.5554e+10 5656.1
- distdryseason          4 2.7948e+09 9.8349e+10 5656.2
- place                  3 3.6293e+09 9.9184e+10 5660.6
- waterusecategory       7 8.0533e+09 1.0361e+11 5665.0
- avepersonsatresidence  1 4.8110e+10 1.4366e+11 5769.9

Step:  AIC=5644.44
unitsTotal ~ place + gender + occupation + waterusecategory + 
    avepersonsatresidence + watersupplytypea + metered_water + 
    distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- watersupplytypea       6 3.3110e+08 9.9446e+10 5633.4
- occupation             2 6.0951e+08 9.9724e+10 5642.2
- gender                 1 6.8207e+05 9.9115e+10 5642.4
- metered_water          1 2.8955e+07 9.9144e+10 5642.5
- distwetseason          3 1.8705e+09 1.0099e+11 5643.8
<none>                                9.9115e+10 5644.4
- distdryseason          4 2.9678e+09 1.0208e+11 5644.8
- place                  3 3.4306e+09 1.0255e+11 5648.1
- waterusecategory       7 8.0933e+09 1.0721e+11 5652.7
+ quality1              11 3.5603e+09 9.5554e+10 5656.1
- avepersonsatresidence  1 5.1014e+10 1.5013e+11 5760.4

Step:  AIC=5633.39
unitsTotal ~ place + gender + occupation + waterusecategory + 
    avepersonsatresidence + metered_water + distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- occupation             2 5.5498e+08 1.0000e+11 5631.0
- gender                 1 7.7698e+05 9.9447e+10 5631.4
- metered_water          1 6.8453e+07 9.9514e+10 5631.6
- distwetseason          3 1.8712e+09 1.0132e+11 5632.7
<none>                                9.9446e+10 5633.4
- distdryseason          4 2.9462e+09 1.0239e+11 5633.7
- place                  3 4.4764e+09 1.0392e+11 5639.9
- waterusecategory       7 8.0997e+09 1.0755e+11 5641.6
+ watersupplytypea       6 3.3110e+08 9.9115e+10 5644.4
+ quality1              11 3.7308e+09 9.5715e+10 5644.5
- avepersonsatresidence  1 5.4280e+10 1.5373e+11 5755.1

Step:  AIC=5630.97
unitsTotal ~ place + gender + waterusecategory + avepersonsatresidence + 
    metered_water + distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- gender                 1 2.8616e+06 1.0000e+11 5629.0
- metered_water          1 1.1051e+08 1.0011e+11 5629.3
- distwetseason          3 1.9705e+09 1.0197e+11 5630.5
<none>                                1.0000e+11 5631.0
- distdryseason          4 3.1100e+09 1.0311e+11 5631.7
+ occupation             2 5.5498e+08 9.9446e+10 5633.4
- place                  3 4.2362e+09 1.0424e+11 5636.8
- waterusecategory       7 7.8261e+09 1.0783e+11 5638.4
+ quality1              11 3.8446e+09 9.6156e+10 5641.8
+ watersupplytypea       6 2.7657e+08 9.9724e+10 5642.2
- avepersonsatresidence  1 5.5669e+10 1.5567e+11 5754.7

Step:  AIC=5628.98
unitsTotal ~ place + waterusecategory + avepersonsatresidence + 
    metered_water + distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- metered_water          1 1.0913e+08 1.0011e+11 5627.3
- distwetseason          3 1.9691e+09 1.0197e+11 5628.5
<none>                                1.0000e+11 5629.0
- distdryseason          4 3.1106e+09 1.0311e+11 5629.7
+ gender                 1 2.8616e+06 1.0000e+11 5631.0
+ occupation             2 5.5706e+08 9.9447e+10 5631.4
- place                  3 4.2347e+09 1.0424e+11 5634.8
- waterusecategory       7 7.9250e+09 1.0793e+11 5636.6
+ quality1              11 3.8274e+09 9.6176e+10 5639.9
+ watersupplytypea       6 2.7671e+08 9.9727e+10 5640.2
- avepersonsatresidence  1 5.6671e+10 1.5667e+11 5754.5

Step:  AIC=5627.29
unitsTotal ~ place + waterusecategory + avepersonsatresidence + 
    distwetseason + distdryseason

                        Df  Sum of Sq        RSS    AIC
- distwetseason          3 1.9960e+09 1.0211e+11 5626.9
<none>                                1.0011e+11 5627.3
- distdryseason          4 3.0837e+09 1.0320e+11 5627.9
+ metered_water          1 1.0913e+08 1.0000e+11 5629.0
+ gender                 1 1.4764e+06 1.0011e+11 5629.3
+ occupation             2 5.9825e+08 9.9514e+10 5629.6
- waterusecategory       7 7.9363e+09 1.0805e+11 5635.0
- place                  3 4.9714e+09 1.0508e+11 5635.1
+ quality1              11 3.9350e+09 9.6178e+10 5637.9
+ watersupplytypea       6 3.6537e+08 9.9747e+10 5638.2
- avepersonsatresidence  1 5.6792e+10 1.5690e+11 5752.9

Step:  AIC=5626.89
unitsTotal ~ place + waterusecategory + avepersonsatresidence + 
    distdryseason

                        Df  Sum of Sq        RSS    AIC
- distdryseason          4 2.2664e+09 1.0438e+11 5625.1
<none>                                1.0211e+11 5626.9
+ distwetseason          3 1.9960e+09 1.0011e+11 5627.3
+ metered_water          1 1.3604e+08 1.0197e+11 5628.5
+ gender                 1 4.1366e+05 1.0211e+11 5628.9
+ occupation             2 7.0399e+08 1.0140e+11 5628.9
- place                  3 3.7414e+09 1.0585e+11 5631.1
- waterusecategory       7 7.9526e+09 1.1006e+11 5634.2
+ quality1              11 4.2831e+09 9.7826e+10 5636.7
+ watersupplytypea       6 3.7931e+08 1.0173e+11 5637.8
- avepersonsatresidence  1 5.6384e+10 1.5849e+11 5749.8

Step:  AIC=5625.13
unitsTotal ~ place + waterusecategory + avepersonsatresidence

                        Df  Sum of Sq        RSS    AIC
<none>                                1.0438e+11 5625.1
+ occupation             2 8.6963e+08 1.0351e+11 5626.8
+ distdryseason          4 2.2664e+09 1.0211e+11 5626.9
+ metered_water          1 7.8002e+07 1.0430e+11 5626.9
+ gender                 1 1.1101e+07 1.0436e+11 5627.1
+ distwetseason          3 1.1787e+09 1.0320e+11 5627.9
- waterusecategory       7 7.2635e+09 1.1164e+11 5630.2
+ quality1              11 4.3929e+09 9.9982e+10 5634.9
+ watersupplytypea       6 2.8177e+08 1.0409e+11 5636.4
- place                  3 7.2098e+09 1.1158e+11 5638.1
- avepersonsatresidence  1 5.6611e+10 1.6099e+11 5746.2

Summary of the fit

r.squared	adj.r.squared	sigma	statistic	p.value	df	logLik	AIC	BIC	deviance	df.residual
0.5935496	0.5285842	19789.3	9.136392	0	40	-3191.005	6464.009	6613.617	95554364430	244

Final fit after stepwise selection.

Final model has been arrived at using stepwise selection whereby the aim is to select those predictors that minimize the AIC.

term	estimate	std.error	statistic	p.value
(Intercept)	24010.0156	5877.87613	4.0848114	0.0000576
placeMutitu Road	-13828.6661	4426.09308	-3.1243505	0.0019683
placeNairobi Road	-1877.0128	4089.61639	-0.4589704	0.6466102
placeWote Road	-5745.8986	4497.93132	-1.2774536	0.2024966
waterusecategoryButcher	-8782.7782	5930.10348	-1.4810497	0.1397137
waterusecategoryHospital	-1678.1723	8430.69042	-0.1990551	0.8423636
waterusecategoryHotel	-10579.0292	5496.46970	-1.9246953	0.0552770
waterusecategoryHouse-hold	-12457.0449	4943.58576	-2.5198400	0.0122953
waterusecategoryOther	-11187.9840	5706.22444	-1.9606632	0.0509060
waterusecategorySchool	-29588.6437	8286.23622	-3.5708183	0.0004185
waterusecategoryShop	-13944.0617	5555.59248	-2.5099144	0.0126384
avepersonsatresidence	165.0039	13.22326	12.4783065	0.0000000
distwetseason0.5-1	-4029.3116	3746.77494	-1.0754080	0.2831146
distwetseason1-2	2402.4401	5080.62464	0.4728631	0.6366779
distwetseason2-5+	1354.2278	5774.58788	0.2345151	0.8147559

Summary of the fit

r.squared	adj.r.squared	sigma	statistic	p.value	df	logLik	AIC	BIC	deviance	df.residual
0.5573857	0.5353338	19313.95	25.27602	0	15	-3333.41	6698.819	6757.865	1.04821e+11	281

A check on the Overall significance or the factor variables.

Wald test:
----------

Chi-squared test:
X2 = 15.3, df = 3, P(> X2) = 0.0016

Wald test:
----------

Chi-squared test:
X2 = 17.9, df = 7, P(> X2) = 0.013

Wald test:
----------

Chi-squared test:
X2 = 181.6, df = 3, P(> X2) = 0.0

Findings from the final model:-

From the regression of Total units of water consumed in the past 3 months on the collected respondent’s information, a list of 4 factors were found to best describe the variability of the amount consumed. These factors include:- Place of residence, Category of water use, Average number of persons needing Water per residence,and Distance(km) from Main water source during the Wet season.

When these factors were considered together they could explain \(\approx\) 53.53% of the variation in the amount of water consumed. For a unit increase in the number of persons needing water in a residence the mean amount of water units consumed increase by a mean of \(167.74\) units holding the other factors constant.

Location was an important predictor of the amount of water units consumed whereby respondents from Kitui road of average consumed more units relative to repondents from Mutitu road, Nairobi road, and Wote road. Category of water usage was also amn important factor that affected units consumed plus the distance the respondent had to cover to the water source during the wet seasons.

Objective 2

To evaluate the potential for management of water demand in Machakos Municipality

Mean Consumption:- a) Overvall and b) By Place;

The tables below shows a summary of the overal water consumption over the 3 months.

Summary of the Overal Mean Water Consumption(??)
Sum	Min	Mean	Median	SD	Max
4433618	0	14680.85	5400	28204.22	210000

The distribution of Water Consumption narrowed down to the “Location/Place” of residence level for the respondents is shown in the table below.

Summary of the Water Consumption(??) By Place of Residence
place	Sum	Min	Mean	Median	SD	Max
Kitui Road	1969140	180	26255.200	9000	38245.374	180000
Mutitu Road	108288	0	1463.351	0	4079.097	27000
Nairobi Road	1071080	0	14093.158	5400	29160.567	210000
Wote Road	1285110	300	16909.342	9600	23836.592	110000

Mean consumption per person!!!.

The idea here is to get the average amount of water consumption or water demand per person. Since information of the average number of persons residing in a given residence was collected we can get each individuals water demand by diving the total units cosumed by the number of residents requiring water in a given residence.

Summary of the Mean Water Consumption(??) per Person
Sum	Min	Mean	Median	SD	Max
451740	0	1510.836	625	4144.65	63000

Summary of the Water Consumption(??) By Place of Residence per Person
place	Sum	Min	Mean	Median	SD	Max
Kitui Road	106548.73	30	1439.8477	900	2217.3012	18000
Mutitu Road	24703.50	0	338.4041	0	741.0370	5000
Nairobi Road	255856.42	0	3366.5318	1800	7554.1422	63000
Wote Road	64631.37	5	850.4128	600	776.1695	4800

Comparing water Supply from the 3 dams with the Mean Consumption(Demand) by the respondent.

Machakos municipality depends on three sources for its water supply, namely Maruba Dam, Nol Turesh supply and boreholes with average production of approximately \(1,300m^3/d\), \(800m^3/d\) and \(120m^3/d\) respectively giving a total production of \(\approx 2,220m^3/d\).

The observation from the residences surveyed is that on average they consumed \(\approx\) 1.46808510^{4} units of water in a duration of three months, where the on average per person demanded \(\approx\) 1510.84 units of water. This in comparison to the water available clearly indicates that the demand by the respondents cannot be fully satisfied by the available water sources.