Agricultural Water Management 60 (2003) 217–226

Economic analysis of rainwater harvesting and

irrigation methods, with an example from China
Tian Yuana,b, Li Fengmina,*, Liu Puhaib
a
The State Key Lab. of Arid Regions Agroecology, Lanzhou University, Lanzhou 730000, PR China
b
Department of Water Resources Engineering, Gansu Agricultural University, Lanzhou 730070, PR China
Accepted 11 November 2002

Abstract

The purpose of this paper is to evaluate the economic feasibility of agriculture with rainwater
harvesting and supplemental irrigation in a semiarid region. The results show the importance of
making full use of every open-air hardened surface to collect rainwater and to establish rainwater
catchment areas by utilizing unoccupied land. The results also show that the usefulness of the
harvested rainwater is enhanced when water saving and prevent seepage techniques are employed.
The results indicate that in order to maximize investment it is essential to select crops with a water
requirement process that coincides with local rainfall events. Potato was found to be the most suitable
crop in the studied region. The economic indices for potato were superior to spring wheat, corn and
wheat/corn intercropping. Therefore, potato production using rainwater harvesting and supplemental
irrigation is the best alternative for cropping systems in the semiarid region of Gansu, China.
# 2002 Elsevier Science B.V. All rights reserved.

Keywords: Rainwater harvesting; Semiarid region; Economic analysis; Supplemental irrigation

1. Introduction

The central part of Gansu province is one of China’s most impoverished regions due to
its harsh climate and unique physical geography. This semiarid region is located in the
transition zone between the monsoon region to the south and the drier region to the north.
The general characteristics of the region are (1) low and an highly variable water resources
with annual rainfall from 316 to 505 mm with large spatial and temporal variability, (2)
large scale soil and water erosion due to porous loess covering the area, and (3) a land-
scape consisting of hills and gullies covering 93% of the total area with slopes of over 158

*
Corresponding author. Tel.: þ86-931-7631675; fax: þ86-931-7668010.
E-mail address: [email protected] (L. Fengmin).

0378-3774/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0378-3774(02)00171-3
218 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226

(Zhao et al., 1995; Zhao, 1996). Investigation on the efficient use of rainwater is critically
important for developing optimum agricultural production systems for this region.
For thousands of years, man has tried to survive in desert regions by skillfully managing
the vital but scarce water resource (Boers and Ben-Asher, 1982). In the central part of
Gansu, China, rainwater harvesting for domestic use has been practiced for thousands of
years. Recently, the collection of rainwater for agricultural use has taken a new direction.
Typically, rainwater harvesting for agriculture encompasses three different systems. The
first system is referred to as Rainwater Conserved Agriculture. This system is a typical rain
fed agricultural system in which the received rainfall is directly converted into stored soil
moisture in the soil profile. The second system, runoff agriculture, is a runoff system that
composes rainwater catchment areas and a command area (Pandey, 1991). In this system,
water is collected in the catchment area and released into fields as needed to irrigate trees or
crops. This method has been given more attentions since 1980s (Frasier, 1980; Boers and
Ben-Asher, 1982; Boers et al., 1986; Ben-Asher and Warrick, 1987; Ojasvi et al., 1999).
The third process is Rainwater Harvesting and Supplemental Irrigated Agriculture
(WHSI). In this system, rainwater catchments are used with one or more reservoirs to
store water, within a command area (Pandey, 1991).
The objective of this paper is to evaluate the economic feasibility of agriculture with
rainwater harvesting and supplemental irrigation in the semiarid regions of Gansu
province, China. Economic analysis was also performed to determine the feasibility of
agricultural production systems for the integrated catchment areas, where water storage
and irrigation equipment are used and special crops such as potato are produced.

2. WHSI components

WHSI is an integrated system that includes rainwater catchment areas, water storage and
transportation methods, a supplemental irrigation method, and a predetermined cropping
pattern.

2.1. Catchment area

The catchment areas used in the study region can be divided into two types. The first
catchment type includes all man-made open-air surfaces, such as highway surfaces, roof
tops and courtyards. The other type also includes man-made surfaces that are designed
especially for rainwater harvesting. The performance of a water harvesting system is
directly related to the relative impermeability of the catchment area. There are various
materials that can be used on the catchment surface to reduce permeability, such as water
repellent, gravel-covered plastic sheeting, concrete, asphalt fiberglass. However, com-
pacted earth is commonly used and is also the cheapest catchment surfaces.

2.2. Water storage

Although reservoirs and storage tanks are used for storing runoff flow, the most
commonly used storage method, in the region, is the dry well or cellar. This is the most
 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226 219

efficient method for storing rainwater from the catchments because it is suitable for
collecting rainwater from small catchment areas. The volume of one cellar is about 30–
40 m3 and one family has three to five cellars. The inner surface of the cellar is generally
covered with a layer of cement.

2.3. Irrigation methods

There are always some distances from the point of water source at the catchment and
reservoir to the field that is ready to irrigate. In the studied region, there are two methods
used to transport the water to the field for irrigation. The simplest method is transportation
with a water tractor in combination with manual watering, the other is transportation with a
water tractor, in combination with a manual driven piston pump for irrigation using a
portable drip irrigation system. The drip system consists of two to five drip tapes about
50 m long, a screen filter, a set of manifold PE pipes, and a piston pump which has a
discharge rate of 1.5–2.5 m3/h and a 13–20 m pressure head.

2.4. Crops

The main crops in this region are spring wheat, potato, oil crops, and spring corn. The
prevailing cropping pattern is spring wheat/corn intercropping in which the corn belt is
80 cm wide and the wheat belt is 100 cm wide. The potato seedlings are planted on ridges
and plant spacing is 25–30 cm.

3. Economic evaluation

3.1. Crop yield and economic profits

In the study region, only 57% of annual rainfall occurs during the growth stage of spring
wheat, therefore crops in the loess plateau region depend, to a great extent, on the
supplemental irrigation period. According to the field experiment, the water requirement of
spring wheat is 357 mm. However, rainfall during the same period is only 288 mm leaving
the region with a total water deficit for all growth stages of 69 mm. The first critical stage of
irrigation occurs in early spring or at the seedling stage since a largest portion of the water
deficit occurs at this time. Irrigation also occurs during the second critical stage when the
crop is heading.
The water use of potato coincides more closely to rainfall events than that of spring
wheat, shown in Fig. 1. Although rainfall does not meet the full water requirement of potato
the value of the deficit is less than that of spring wheat. The critical stage of potato irrigation
occurs between the last 10 days of May and the first 10 days of June. Limited water from
rainwater collection can be applied by manual or portable drip irrigation methods as
required by the crops during these critical stages of crop growth.
The investigation was carried out near Dingxi county, which is representative of the
central part of Gansu. A large amount of the area in the county is gullies and hills, which
account for 88% of the total area. The altitude is 1900–2250 m, annual rainfall is 425 mm,
220 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226

Fig. 1. Comparison between rainfall process and water requirement of spring wheat and potato, Dingxi.

and annual pan evaporation is 1510 mm. Since the 1990’s, more than three water cellars per
family have been built to store rainwater collected from various catchment areas. The type
of catchment areas include roof tops, concrete courtyard, highway surfaces, and other
manmade surfaces. Reservoir for water storage are always cellars lined with cement.
Two types of farming systems, one is traditional rain fed farming, and the other is
supplemental irrigation farming with rainwater harvesting, were chosen to investigate. The
yields of four crops and planting pattern, including spring wheat, corn, potato, and wheat/
corn inter-cropping, at the rain fed farming and supplemental irrigation farming, were
surveyed, respectively. Meanwhile, the cost of supplemental irrigation for the crops was
investigated.
The average yield at the rain fed farming and the average increase in yield and output
value at the supplemental irrigation farming, based on the investigated data and market
price for the 3 years, are shown in Table 1.

Table 1
Response of crop yield to rainwater harvesting and supplemental irrigation

Crop and Average yield Irrigation Increased Unit price Total increased output
planting pattern on rain fed (kg/ha) (m3/ha) yield (kg/ha) (RMB/kg)a value (RMB/ha)

Spring wheat 2431 600 1410 1.10 1551

Corn 2953 750 2622 0.80 2098
Wheat/corn 3793 666 2390b 0.92c 2199
inter-cropping
Potato 15000 600 7500 0.40 3000
a
Conversion rate of US dollar to Renminbi (RMB) is about US$ 1 ¼ 8.3 RMB.
b
Composite increased yield including spring wheat and corn.
c
Composite unit price including spring wheat and corn.
 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226 221

3.2. Added costs

Catchment area, water storage and irrigation facilities are three basic factors that impact
on the cost of rainwater harvesting and supplemental irrigation agriculture. Total cost of
supplemental irrigation is a function of catchment area, tank geometry, rainfall pattern,
evaporation, land management factors, seepage rates of the catchment and the tank, and
irrigation depth and area (Pandey, 1991). The cost also relates to irrigation methods, crops,
and planting models (Tiwari et al., 1997). In this paper, these added costs are included when
common dry land agriculture is compared with supplemental irrigation.
A four phase cropping experiment was conducted at Dingxi (Zhao, 1996) to obtain the
relative best yields for spring wheat, corn, potato and wheat/corn using the minimum
supplemental irrigation level. The required catchment area and total number of cellars
was then set based on standardized size and type utilized in the region (Table 2). Irrigation
methods employed at the critical growth stage include portable drip or micro spray
irrigation systems and manual watering point or by individual furrow. The cost of the
supplemental irrigation method including the added cost of the irrigation system was
compared with the traditional rain fed farming system and runoff agriculture system
(Table 3).

Table 2
The required catchment area and cellar number under minimum irrigation

Crop and Total irrigation Average runoff Required catchment Required cellar
planting pattern (m3/ha) efficiencya (m2/ha) (m3/ha)

Spring wheat 600 0.79 (0.23) 1765 (6000) 15

Corn 750 0.79 (0.23) 2206 (7500) 19
Wheat/corn inter-cropping 666 0.79 (0.23) 1959 (6660) 17
Potato 600 0.79 (0.23) 1765 (6000) 15
a
The values not in brackets pertain to concrete and asphalt road surface catchment area, while the values
inside parentheses pertain to compacted earth catchment.

Table 3
Added cost of supplemental irrigation with rainwater (RMB/ha)a

Crops Catchmentb Water cellar Drip irrigation Operation costc

Spring wheat 265 (1800) 6000 600 150 (300)

Corn 330 (2250) 7600 600 150 (300)
Wheat/corn inter-cropping 294 (1998) 6800 600 150 (300)
Potato 265 (1800) 6000 600 150 (300)
a
Conversion rate of US dollar to Renminbi (RMB) is about US$ 1 ¼ 8.3 RMB.
b
The values not in brackets pertain to concrete and asphalt road surface catchment area, while the values
inside brackets pertain to compacted earth catchment.
c
The values not in brackets pertain to manual irrigation, while the values in brackets pertain to irrigation
with a portable drip system.
222 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226

3.3. Economic evaluation

Two dynamic economic indices, financial net present value (NPV) and financial internal
rate of return (IRR), are used to compare each scheme of rainwater harvesting and
supplemental irrigation. It is assumed that the service life of catchment and water cellar is
10 years and the portable drip irrigation systems have a useful life of 5 years. Analysis is
preformed over a 10-year period with renewal costs considered for capital items with a
useful life less then 10 years. A discount rate of 6% based on the code of water economic
calculation (SD139-85) is used for NPV calculations. During the analysis period, profits
and operational costs are constant. Therefore, the present value of gross profits (GB) and
total costs (TC) are, respectively,
X n
Bt
GB ¼ (1)
t¼1 ð1 þ iÞt
X
m
K3t Xn
Ct
TC ¼ K1 þ K2 þ t þ t (2)
t¼0 ð1 þ iÞ t¼1 ð1 þ iÞ

where Bt is the gross profit at the tth year, Ct the cost at the tth year, i the discount rate, K1
the investment on catchment areas, K2 the investment on water cellars, K3t the investments
on irrigation equipment at the tth year, m the replacement time of irrigation equipment, and
n the calculation period.
Therefore, the net present value of profit is
NPV ¼ GB  TC (3)
If NPV > 0, the scenario is accepted, if not, the scenario is unfeasible. The duration of
time when the net revenue compensates for the total investment is the capital recovery
period.
In order to calculate the internal rate of return the solution of IRR in the Eq. (4) must be
found, that is
" #
Xn
Bt Xm
K3t Xn
Ct
t  K1 þ K2 þ t þ t ¼ 0 (4)
t¼1 ð1 þ IRRÞ t¼0 ð1 þ IRRÞ t¼1 ð1 þ IRRÞ

where IRR is the internal rate of return. The IRR is acceptable if it is greater than minimum
expected interest rate.
According to the Table 3 and Eqs. (1)–(3), the financial net present values for the four
crops under the four treatment phases are calculated.
Figs. 2 and 3 represent the comparison between two kinds of catchments, concrete or
asphalt road surface and compacted original earth surface, under the portable drip
irrigation system. The NPV’s of the compacted original earth catchment are obviously
less than that of the concrete or asphalt catchments. It is also evident that in four cropping
patterns the NPVof potato is the highest and spring wheat is the lowest. When the concrete
catchment and portable drip irrigation are used, the capital recovery period of potato is <3
years and more than 6 years for wheat. Similarly, when the compacted original earth
catchment and portable drip irrigation are used, the capital recovery period of potato is
slightly more than 3 years and wheat is approximately 8 years.
 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226 223

Fig. 2. Comparison of four crops’ NPV at catchment of concrete and asphalt road surface and portable drip
irrigation (interest rate ¼ 6%).

Figs. 4 and 5 show the comparison between two catchments of concrete or asphalt road
surface and compacted original earth surface under the condition of manual furrow
irrigation. The tendency of NPV is similar to the situation of the portable drip irrigation
scenario. However, the capital recovery period of potato is slightly less than the portable
drip irrigation system.
The two indices, financial NPVand capital recovery period show that in order to increase
crop yield and economic return it is important to make full use of all existing impermeable

Fig. 3. Comparison of four crops’ NPV at catchment of compacted original earth surface and portable drip
irrigation (interest rate ¼ 6%).
224 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226

Fig. 4. Comparison of four crops’ NPV at catchment of concrete and asphalt road surface and manual irrigation
(interest rate ¼ 6%).

and open-air surfaces for rainwater harvesting. It is also important to select crops with water
requirement processes that coincide with prevailing rainfall events. The NPVs of potato
show that potato has the best return on investment other crops and the capital recovery
periods of potato are less than that of others no matter what kind of catchment and irrigation.
The comparison of financial internal rate of return for all treatments is shown in Fig. 6.
The IRR of concrete catchment is higher than that of earth catchment when the same

Fig. 5. Comparison of four crops’ NPV at catchment of compacted earth surface and manual irrigation (interest
rate ¼ 6%).
 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226 225

Fig. 6. Comparison of internal rate of return at four treatments. E-Drip: earth catchment and drip irrigation;
E-Manual: earth catchment and manual watering; C-Drip: concrete or asphalt catchment and drip irrigation; and
C-Manual: concrete or asphalt catchment and manual watering.

irrigation method is used. The IRR of different irrigation methods does not have distinct
differences. However, the IRR of potato is distinctively higher than other crops with the
IRR values for the four potato treatments as high as 31.09–41.78%. In this region, it is not
economically feasible to grow spring wheat and corn using the earth catchment system
regardless of irrigation type. The IRR of spring wheat/corn intercropping is 22.49 and
23.53% for drip and manual irrigation, respectively, when there is a compatible concrete or
asphalt surface for rainwater harvesting. Potato is the most suitable crop for using the
WHSI farming system.

4. Conclusions

In the central part of Gansu, rainwater harvesting and supplemental irrigation farming
has an advantage over traditional rain fed and runoff farming in increasing crop yield of
spring wheat, corn, and potato. The WHSI farming system emphasizes irrigation at the
critical stage when the largest deficit between water requirement and rainfall occurs.
Therefore, to obtain most economic return from WHSI, it is necessary to choose suitable
crops and growing patterns as well as catchment types and irrigation methods.
The results of the economic analysis show that in semiarid regions all open-air hardened
surfaces such as, highway surfaces, courtyards and roofs should be used to collect
rainwater since it is the cheapest way of rainwater harvesting. In addition, unoccupied
land should be used to establish rainwater catchment areas to help supplement the harvest
of rainwater. The adoption of other conservation techniques including portable micro-
irrigation and prevent seepage techniques should be considered to maximize the usefulness
of the harvested rainwater. The most important decision is to choose crops with water
226 T. Yuan et al. / Agricultural Water Management 60 (2003) 217–226

requirement processes that best coincide with prevailing local rainfall events. This
investigation shows that potato is the crop best suited for the region based on both
economic indices NPV and IRR and its ability to utilize available rainwater efficiently. The
NPV and IRR of potato are much greater than that of spring wheat and corm. In addition,
potato had the best response to supplemental irrigation compared to spring wheat, corn, and
the spring wheat and corn intercropping. In the Loess plateau, especially in Gansu center,
there are many favorable conditions for growing potato, such as cool climate, loose soil,
and a long history and rich experience of planting the crop. Therefore, develop potato
production using WHSI provides the best investment opportunity.

References

Ben-Asher, J., Warrick, A.W., 1987. Effect of variations in soil properties and precipitation on microcatchment
water balance. Agric. Water Manage. 12, 177–194.
Boers, T.M., Ben-Asher, J., 1982. A review of rainwater harvesting. Agric. Water Manage. 5, 145–158.
Boers, Th.M., Zondervan, K., Ben-Asher, J., 1986. Micro-catchment-water-harvesting (MCWH) for arid zone
development. Agric. Water Manage. 12, 21–39.
Frasier, G.W., 1980. Harvesting water for agricultural, wildlife and domestic uses. J. Soil Water Conserv. 35 (3),
125–128.
Ojasvi, P.R., Goyal, R.K., Gupta, J.P., 1999. The micro-catchment water harvesting technique for the plantation
of jujube (Zizypphus mauritiana) in an agroforestry system under arid conditions. Agric. Water Manage. 41,
139–147.
Pandey, S., 1991. The economics of water harvesting and supplemental irrigation in the semiarid tropics of India.
Agric. Syst. 36, 207–220.
Tiwari, K.N., Reddy, K.Y., 1997. Economic analysis of trickle irrigation system considering planting geometry.
Agric. Water Manage., 34:195–206.
Zhao, S.L., 1996. Introduction of Catchment Agriculture. Publishing House of Science & technology, Shaanxi.
Zhao, S.L., Li, F.M., Wang, J., 1995. On the development of water harvesting agriculture in the semiarid area
northwest China. Acta Bot. Boreal. Occident. Sin. 15 (8), 9–13.