A Study on the Coordinated Operation of Reservoirs with Low Watershed Magnification Ratios Using Surplus Storage Capacity

Abstract

This study proposes a hardware-based approach to address agricultural water shortages by directly improving water supply operations, rather than estimating agricultural water demand or supply. Unlike previous studies that focus on evaluating water supply capacity or predicting reservoir inflows through modeling or data-driven methods, this work proposes an operational strategy involving the physical interconnection of reservoirs. Specifically, the study investigates the coordinated use of surplus storage capacity from reservoirs with high watershed ratios to support those with low watershed ratios, thereby enhancing overall water supply reliability. Reservoir inflows were estimated using the Hydrological Operation Model for Water Resources Systems (HOMWRS). The analysis was conducted on reservoirs managed by the Korea Rural Community Corporation (KRC), selected based on data accessibility and availability.

1. Introduction

Recent climate change has caused increasingly uneven spatial distribution of precipitation, leading to localized droughts. Although the nationwide annual precipitation remains close to the long-term average, drought damage has frequently occurred in many areas, particularly in regions cultivating upland crops. Areas such as western Chungcheongnam-do and southern Gyeonggi-do have been repeatedly affected, with the 2016 extreme drought in Anseong, Gyeonggi Province—estimated to have a 200-year return period—serving as a notable example. In the past, drought impacts were primarily driven by insufficient rainfall. However, recent improvements in irrigation infrastructure, supported by agricultural development projects, have enhanced the overall drought resilience of many regions [1]. Nevertheless, spring droughts remain a significant challenge, as they occur before the onset of concentrated summer rainfall, and current measures fall short in securing water resources and balancing regional supply and demand. Therefore, there is an urgent need for more efficient water utilization strategies and improved water supply measures for reservoirs with low watershed ratios.

Reservoir inflows consist of upstream, lateral, and tributary components [2]. However, in regions with insufficient meteorological data, estimating inflows remains challenging. Various methods have been employed to estimate reservoir inflows, utilizing both physical-based and data-driven approaches [3,4,5]. Earlier studies predominantly relied on physical-based models, whereas recent research has increasingly focused on data-driven models for inflow prediction [6]. For example, one study compared three approaches for reservoir inflow prediction [7]: the TANK model, a data-driven AI model combining satellite data with recurrent neural networks, the Parsimonious Eco-Hydrologic Model (PEHM), and decision tree-based algorithms. These studies demonstrate that reservoir inflow prediction has been explored using diverse methodologies [8,9,10]. Although many methods have been applied to secure reservoir inflows across numerous sites, limitations in prediction accuracy and lack of sufficient data continue to constrain research efforts [11,12]. In practice, due to physical limitations, alternative hardware-based solutions such as the development of underground wells and installation of subsurface dams in riverbeds have been implemented to store groundwater and enhance drought resilience. Concurrently, efforts have been made to resolve regional imbalances in water supply and demand. One such approach involves redistributing surplus storage from reservoirs with sufficient capacity to regions with low watershed ratios and limited water supply capabilities [13]. Nevertheless, reservoirs with inherently low watershed ratios lack the capacity to reliably supply water, making them structurally vulnerable under drought conditions [14].

The design of reservoir water supply systems is typically based on projected downstream water demand prior to construction, with the required storage capacity determined accordingly [15]. Studies have assessed the impacts of climate change scenarios on both reservoir inflows and downstream demand [16]. In Korea, efforts have been made to more accurately evaluate water supply capacity by applying a 95% reliability criterion to multipurpose dams and river basins, particularly in the Nakdong River watershed [17]. More recently, numerous studies have examined the performance of agricultural water supply systems under various climate change scenarios. In response to increasing water demand, several studies have evaluated reservoir water supply capacity and proposed improved operational strategies. For instance, some research has demonstrated that adjusting downstream release volumes and target operating water levels can significantly improve agricultural water supply [18]. Other studies have focused on estimating the supply capacity of agricultural reservoirs, for example, employing the DIROM (Daily Irrigation Reservoir Operation Model) to estimate supply by adjusting release coefficients [19]. More recent work has utilized ICT-based systems and automated measurement data to estimate and analyze water supply from agricultural reservoirs [20], reporting unexpected results that diverged from the typical inverse relationship between water demand and rainfall.

Most prior studies have focused on estimating agricultural water supply through evaluating supply capacity and modeling-based inflow prediction. However, with advances in technology, there is an increasing need for research that directly addresses water shortages through hardware-based operational strategies, rather than relying solely on supply estimation. This study therefore examines the potential to enhance water supply capacity by reallocating surplus storage from reservoirs with high watershed ratios and sufficient supply to those with low watershed ratios and limited capacity. In addition, construction costs were approximately estimated based on the connection method and distance for each reservoir, and both the economic feasibility and practical applicability were evaluated. Reservoir inflows were estimated using the Hydrological Operation Model for Water Resources Systems (HOMWRS). The analysis targeted reservoirs managed by the Korea Rural Community Corporation (KRC), selected for their data accessibility. Hydrological analysis was conducted on reservoirs identified as potential suppliers or recipients based on criteria such as watershed ratio and geographical proximity.

2. Study Area and Materials

2.1. Study Area

Of the 17,313 reservoirs distributed across the Republic of Korea, this study focused on 229 reservoirs located in the Chungcheongnam-do region, which are managed by the Korea Rural Community Corporation (KRC. Naju, republic of korea). Reservoir basin ratio is the basin area divided by the benefit area. From these, a subset of reservoirs suitable for hydrological interconnection was identified as the study targets. The selection criteria included a watershed ratio of 2 or less indicating insufficient natural inflow and a storage capacity of at least 300,000 m³ to ensure a minimum operational scale. The nationwide distribution of irrigation-related facilities is summarized in

Table 1. Current status of irrigation infrastructure.

Item	Total		KRC Management		Local Government Management
	Number	Irrigation Area (ha)	Number	Irrigation Area (ha)	Number	Irrigation Area (ha)
Total	72,786	727,905	13,911	498,785	58,875	229,091
Reservoir (Chungnam)	17,313 (953)	440,281	3403 (229)	337,561 (52,695)	13,910 (724)	102,720
Pumping Station	8233	181,947	4473	153,606	3760	28,341
Sea dike, etc.	47,240	105,677	6035	7618	41,205	98,030

. In the Chungcheongnam-do region alone, the 229 reservoirs collectively irrigate approximately 52,695 hectares.

From the 229 reservoirs in the Chungcheongnam-do region, 22 were selected for analysis according to criteria indicating suitability for hydrological interconnection—specifically, a storage capacity of at least 300,000 m³ and a watershed ratio of 2 or less. Among these, five reservoirs that could feasibly receive water from reservoirs with higher watershed ratios were chosen for detailed investigation. The characteristics and locations of these five reservoirs are presented in

Table 2. Characteristics of reservoirs below W:Basin Area/I:Irrigation Area ratio 2.0 in Chungnam.

Number	Item	Location (Coordinate: WGS84)	Basin Area (ha)	Irrigation Area (ha)	Watershed Magnification Ratio	Storage (103 m3)
Total	5		6537	4475	1.70	20,838
1	Yugye	36.525967, 127.163951	190	95	2.00	748
2	Okgye	36.423203, 126.637995	220	115	1.91	904
3	Dongbu	36.126876, 126.775670	2834	2477	1.14	10,734
4	Songak	36.879810, 126.695988	254	161	1.58	463
5	Seobu	36.126431, 126.690561	3039	1627	1.87	7989

, while their spatial distribution—distinguishing between supplying (high watershed ratio) and receiving (low watershed ratio) reservoirs—is illustrated in Figure 1. Figure 2 presents the land cover and land use maps of the study site.

2.2. Hydrological and Meteorological Data the Materials

The reservoirs used in this study have relatively small catchment areas and therefore lack dedicated hydrological and meteorological monitoring stations. Due to the difficulty in obtaining long-term observational data, weather and hydrological records from nearby observation stations were used. To select representative stations for each reservoir, stations located within the Thiessen polygon of the reservoir and exhibiting similar topographical characteristics were identified and adopted for use (

Table 3. Meteorological station name and location.

Item	Observatory	Location (Coordinate: WGS84)	Data Period	Observation Year
Yugye	Daejeon	36.37199, 127.3721	1971~2017	47 Year
Wolgok
Okgye	Boryeong	36.32724, 126.55744	1973~2017	45 Year
Cheongcheon
Dongbu	Gunsan	36.0053, 126.76135	1968~2017	50 Year
Munsan
Songak	Cheonan	36.76217, 127.29282	1973~2017	45 Year
Obong
Seobu	Gunsan	36.0053, 126.76135	1968~2017	50 Year
Jongcheon

).

The meteorological data used in this study were obtained from the Daejeon, Boryeong, Gunsan, and Cheonan weather stations, with records spanning up to 50 years (from 1968 to 2017). Daily data were collected for variables relevant to water balance analysis, including temperature, precipitation, number of rainy days, evaporation, and wind speed (Figure 3).

3. Research Method

3.1. Surplus Water Analysis Method

In this study, surplus water is defined as the overflow discharged from a reservoir when the water level exceeds its full supply level. Such overflow, released into nearby rivers without being used for agricultural purposes, is considered ineffective. To estimate the surplus volume, a water balance analysis was performed, focusing on two key variables: reservoir inflow and agricultural water demand. For this purpose, the Hydrological Operation Model for Water Resources Systems (HOMWRS), a reservoir water balance model developed for ungauged catchments, was applied. In this study, HOMWRS, a water facility simulation system developed by the Korea Rural Community Corporation, was utilized for the analysis of agricultural water balance [21].

Input data included existing reservoir information (e.g., location, irrigated area, watershed area), data from representative weather stations, cropping types and schedules, and crop coefficients. The surplus water volume estimated by the model was then reallocated by linking reservoirs with high watershed ratios to those with low watershed ratios, enabling an assessment of potential improvements in water supply capacity at the target reservoirs.

3.2. Reservoir Water Supply Capacity Method

Daily and decadal (10-day interval) inflows to the reservoirs were estimated using the inflow estimation module of the HOMWRS model. The DIROM model was applied for daily inflow estimation, enabling application to ungauged catchments, while decadal inflows were calculated using the Kajiyama method [22]. The DIROM model, a modified version of Sugawara’s TANK model tailored to the hydrological characteristics of irrigation reservoir watersheds in Korea [23], was employed to simulate the daily inflow of each target reservoir. For irrigation water demand estimation, daily paddy field requirements were calculated using the Penman equation [24], and decadal requirements were determined using the Blaney–Criddle method [25]. Daily upland crop water demand was estimated using the Penman–Monteith method [26]. Water balance analysis was conducted by applying inflows derived from the DIROM model and demands calculated using the modified Penman method. For decadal water balance calculations, inflows from the Kajiyama method and demands from the Blaney–Criddle method were used.

Evapotranspiration was estimated using the Penman method, and effective rainfall was calculated based on a ponding depth threshold of 60 mm. Although rice cultivation is generally divided into three growth stages—nursery, transplanting, and main field growing—regional variations in the timing of these stages exist. Therefore, cropping schedules were classified by region (southern and central zones) and these variations were incorporated into the irrigation demand estimation. Apart from cropping schedules, key parameters for estimating irrigation water requirements included infiltration rate, channel loss, and water depth. The input values used for each interconnected reservoir are summarized in

Table 4. Parameters for estimating irrigation water requirements.

Item	Irrigation Area (ha)	Infiltration (mm/day)	Channel Loss (%)	Water Depth (mm)
Yugye	95	5.7	10.0	80~20
Wolgok	36	5.7	10.0	80~20
Okgye	115	3.9	10.0	80~20
Cheongcheon	2638	5.0	10.0	80~20
Dongbu	2477	6.0	10.0	80~20
Munsan	261	6.0	10.0	80~20
Songak	161	4.0	10.0	80~20
Obong	367	4.1	10.0	80~20
Seobu	1627	5.6	10.0	80~20
Jongcheon	385	4.4	10.0	80~20

.

For the purpose of water supply, the distances between reservoirs were investigated as follows: Yugye–Wolgok, 1.7 km; Okgye–Cheongcheon, 2.8 km; Dongbu–Munsan, 2.5 km; Songak–Obong, 3.5 km; and Seobu–Jongcheon, 3.5 km. All reservoir pairs were located within a 4 km range. Among them, the Dongbu and Munsan Reservoirs could be interconnected via open channels, whereas the remaining pairs would require linkage through irrigation systems.

3.3. WaterBalance Method

For the water balance analysis of the target reservoirs, the following equation was applied to estimate daily changes in storage volume, using model-estimated inflow and water demand data:

St_n = St_n−1 + It + Ut + Pt − (Rt + Ot + Et + Gt + Dt)

where

• St_n: daily change in storage volume;
• St_n−1: storage volume on the previous day;
• It: inflow;
• Ut: groundwater inflow;
• Pt: precipitation on the reservoir surface;
• Rt: release (irrigation, environmental flow, or industrial use);
• Ot: overflow (surplus water discharge);
• Et: evaporation from the reservoir surface;
• Gt: percolation loss through the reservoir bed;
• Dt: seepage loss through the reservoir embankment.

Among these, the most critical factors in the reservoir water balance analysis are inflow, irrigation demand, and release volume. Releases may include water for irrigation, environmental flow maintenance, or industrial use. On the other hand, groundwater inflow, percolation losses, and seepage through the embankment were found to have negligible impacts on the overall water balance.

Based on the results of the water balance analysis, this study evaluated water supply capacity using an annual reliability index—one of the standard indicators for assessing supply performance. The following equation was used to calculate water supply reliability:

R = (1 − Ts/Tt) × 100

where

• R: water supply reliability (%);
• Ts: number of years with water shortages;
• Tt: total number of years in the analysis period.

4. Results

4.1. Inflow

Inflow volumes to the reservoir catchments were analyzed by considering land cover and watershed conditions. The land cover status was analyzed based on the proportion of rice paddies, dry fields, and forests, using a land cover map. For the Wolgok Reservoir, the average annual inflow over the 47-year analysis period was 1988.9 × 10³ m³, with a minimum of 648 × 10³ m³ recorded in 2015 and a maximum of 2359.1 × 10³ m³ in 1998. The difference between the minimum and maximum annual inflows for both Yukye and Wolgok reservoirs exceeded 2000 × 10³ m³, indicating substantial interannual variability, see

Table 5. Inflows to Yukye and Wolgok reservoirs.

Year (47 Year)	Rainfall (mm)	Yukye Reservoir Inflow		Wolgok Reservoir Inflow
		(mm)	(103 m3)	(mm)	(103 m3)
Average	1339.6	710.1	1349.2	786.1	1988.9

.

For the Okgye Reservoir, the average annual inflow over the 45-year period was 1596.6 × 10³ m³, with a minimum of 794.8 × 10³ m³ in 1973 and a maximum of 2820.8 × 10³ m³ in 1987. In contrast, the Cheongcheon Reservoir recorded a significantly higher average annual inflow of 57,393.2 × 10³ m³, approximately 36 times greater than that of Okgye. Due to its large catchment and inflow volume, Cheongcheon also showed a wide variation between minimum and maximum annual inflows. The runoff ratio for Cheongcheon was estimated to be approximately 60%, see

Table 6. Inflows to Okgye and Cheongcheon reservoirs.

Year (45 Year)	Rainfall (mm)	Okgye Reservoir Inflow		Cheongcheon Reservoir Inflow
		(mm)	(103 m3)	(mm)	(103 m3)
Average	1205.2	725.7	1596.6	818.7	57,393.2

.

For Dongbu and Munsan Reservoirs, the average annual inflow over the 50-year period was 21,410.2 × 10³ m³. The minimum inflow occurred in 1988 at 11,245.2 × 10³ m³, while the maximum was observed in 1987 at 34,978.7 × 10³ m³. On average, Dongbu Reservoir received approximately three times more inflow than Munsan Reservoir. The estimated runoff ratio for Dongbu was 62.5%, which was relatively higher than that of the Wolgok Reservoir, see

Table 7. Inflows to Dongbu and Munsan reservoirs.

Year (50 Year)	Rainfall (mm)	Dongbu Reservoir Inflow		Munsan Reservoir Inflow
		(mm)	(103 m3)	(mm)	(103 m3)
Average	1209.1	755.5	21,410.2	686.7	6592.5

.

For Songak and Obong Reservoirs, the average annual inflow over the 45-year analysis period was 1632.2 × 10³ m³. The minimum inflow was recorded in 1988 at 807.4 × 10³ m³, while the maximum occurred in 2011 at 2777.9 × 10³ m³. On average, Obong Reservoir received approximately nine times more inflow than Songak Reservoir. In addition, the inflow to Obong Reservoir showed a significantly wider range between minimum and maximum annual values, see

Table 8. Inflows to Songak and Obong reservoirs.

Year (45 Year)	Rainfall (mm)	Songak Reservoir Inflow		Obong Reservoir Inflow
		(mm)	(103 m3)	(mm)	(103 m3)
Average	1223.3	642.6	1632.2	729.3	14,732.7

.

For Seobu and Jongcheon Reservoirs, the average annual inflow over the 47-year analysis period was 23,189.7 × 10³ m³. The minimum inflow was recorded in 1988 at 12,189.4 × 10³ m³, while the maximum occurred in 2011 at 34,634.5 × 10³ m³. Jongcheon Reservoir, which is hydraulically connected to Seobu, also recorded its minimum inflow in 1988. However, its maximum inflow occurred in 2000, differing from Seobu Reservoir, see

Table 9. Inflows to Seobu and Jongcheon reservoirs.

Year (47 Year)	Rainfall (mm)	Seobu Reservoir Inflow		Jongcheon Reservoir Inflow
		(mm)	(103 m3)	(mm)	(103 m3)
Average	1209.1	763.1	23,189.7	735.3	8823.7

.

A comparison of watershed inflows across the five target reservoirs revealed that inflow volumes varied according to watershed size and regional differences in rainfall. Moreover, the variability of inflows differed among the reservoirs, with notable disparities between maximum and minimum annual inflows.

4.2. Water Requirement

Irrigation water requirements for the study sites were estimated by accounting for evapotranspiration, percolation, and effective rainfall. The average values for each reservoir are summarized in

Table 10. Irrigation water requirements by reservoir. (PET Potential Evapotranspiration, AET: Actual Evapotranspiration, NIR: Net Irrigation Requirement, GIR: Gross Irrigation Requirement).

Reservoir Name	PET (mm)	AET (mm)	Water Consumpiton (mm)	Rainfall (mm)	Effective Rainfall (mm)	NIR (mm)	GIR (mm)	GIR (103 m3)
Yugye	665	709.2	1736.0	977.3	671.1	736.4	818.3	777.4
Wolgok	665	709.2	1777.5	977.3	674.6	762.8	847.6	305.1
Okgye	649.6	705.6	1777.3	867	607.3	835.4	928.2	1067.4
Cheongcheon	649.6	705.6	1777.3	867	607.3	835.4	928.2	24,485.6
Dongbu	676.1	732.5	1580.1	850.8	585	712.8	791.9	19,616.6
Munsan	676.1	732.5	1538.6	850.8	580.3	685.9	762.1	1989.2
Songak	666	724.6	1497.3	875	582.5	652.4	724.9	1167.1
Obong	666	724.6	1786.1	875	615.6	839.8	933.1	3424.5
Seobu	676.1	732.5	1746.2	850.8	602.3	820.7	911.9	14,836.2
Jongcheon	676.1	732.5	1580.1	850.8	585	712.8	791.9	3049

.

At Yugye Reservoir, the average consumptive use over the 47-year period was 1736 mm. Of this, 671.1 mm was supplied by effective rainfall, resulting in a gross irrigation requirement of 818.3 mm. A comparison of annual irrigation volumes showed a 2.7-fold difference between a wet year (2003: 424 × 10³ m³) and a dry year (2015: 1152.9 × 10³ m³). During wet years, rainfall can be directly utilized for crop water needs, reducing the reliance on irrigation. In contrast, limited rainfall during dry years leads to a substantial increase in irrigation demand.

At Wolgok Reservoir, the average consumptive use was 1777 mm, with 674.6 mm provided by effective rainfall and a gross irrigation requirement of 847.6 mm.

Similar patterns were observed across the remaining eight reservoirs: gross irrigation requirements varied significantly between wet and dry years, reflecting the strong influence of interannual rainfall variability on reservoir-supplied irrigation demand.

4.3. Water Balance and Supply Capacity

At Yugye Reservoir, a daily water balance analysis over a 47-year period revealed that the reservoir water level dropped to the dead low water level in four years (1994, 1995, 2015, and 2016), resulting in a water supply reliability of 91.5%. The maximum annual water shortage was recorded in 1995, reaching 373 × 10³ m³.

Table 11. Water supply reliabilities by reservoir. (FWL: Full Water Level).

Reservoir Name	FWL (EL.m)	Reservoir Water Shortage (103 m3)	No. of Occurrence Year with Water Shortage	Water Supply Reliability (%)
Yugye	81.4	747.74	4	91.5
Wolgok	73.5	175.24	0	100
Okgye	104.2	828.7	8	82.2
Cheongcheon	40.2	20,753.1	2	95.6
Dongbu	15.5	10,733.8	20	60.0
Munsan	11.4	973.2	11	78.0
Songak	32.9	463.0	32	28.9
Obong	10.1	875.0	27	40.0
Seobu	21.16	7989.0	14	72.0
Jongcheon	42.9	3353.1	0	100

summarizes the water supply reliability for all ten analyzed reservoirs. Notably, Songak and Obong Reservoirs, which were constructed before the establishment of modern design standards, exhibited very low reliability values.

Table 12. Result of water balance of each reservoir by connecting. (Min WL: Minimum Water Level).

Connect Reservoir	Item	Inflow (103 m3)	Outflow (103 m3)	Water Balance	Max. Required Storage	Water Shortage	Min. WL (EL.m)	Min. Storage	Overflow (103 m3)
Yugye	Mean	3115.8	777.4	2338.4	329.2	0	76.66	418.6	2338.4
	Max.	5697.7	1195.2	4993.3	731.6	0	79.80	615.2	4993.3
	Min.	1215.3	424	62.4	132.5	0	66.62	16.2	382.4
Okgye	Mean	36,169.1	1067.4	35,101.7	567.9	37.9	97.11	275.8	35,130.8
	Max.	89,629.5	1438.1	88,909.6	1215.5	417.3	102.2	596.2	88,909.6
	Min.	3967.9	717.5	3002.9	202	0	91.00	0	2512.1
Dongbu	Mean	27,557	19,616.6	7940.4	13,584.6	4278	9.68	2063.6	10,226.6
	Max.	47,168.2	27,908.3	33,533.3	30,314.6	19,537.1	13.34	6332.8	32,354.2
	Min.	12,787.8	11,500.2	−15,121	4444.8	0	7.00	0	0
Songak	Mean	13,564.5	1167.1	12,397.4	556.8	126	27.32	35	12,526
	Max.	25,407.7	1613	24,817.7	1281	815.2	30.94	273.1	24,818
	Min.	3900.6	453.9	2287.6	192.7	0	26.58	0	3102.8
Seobu	Mean	29,901.8	14,836.2	15,065.6	7602	1685.2	14.33	1714	16,385
	Max.	52,026.3	20,483.4	41,481.2	16,715.2	9366.9	18.93	4576	41,481
	Min.	13,514.4	9312.1	−6968.9	2772.7	0	10.00	0	2627.1

presents the results of the annual average water balance analysis for the five target reservoirs after interconnection.

Hydrological analysis showed that interconnecting the reservoirs improved overall water supply reliability, although the level of improvement differed by site. At Yugye Reservoir, reliability increased from 91.5% to 100%, with no water shortages occurring after the reservoirs were connected. Dongbu and Songak Reservoirs also showed improvements of approximately 10% and 7%, respectively. Given Dongbu’s large service area (2477 ha), the interconnection was deemed particularly effective.

At Okgye Reservoir, the reliability increased by only 2.2%. This limited improvement is attributed to its insufficient storage capacity, which prevented it from capturing the surplus discharge from Cheongcheon Reservoir. Similarly, although Seobu Reservoir has a relatively large effective storage volume (7989 × 10³ m³), the impact of interconnection was minimal due to the small volume of surplus inflow from Jongcheon Reservoir.

5. Conclusions

This study evaluated the interconnection of ten reservoirs in Chungcheongnam-do, selected based on watershed ratio and geographic proximity, to assess the potential of surplus water redistribution for improving water supply capacity. By comparing water supply reliability and shortage volumes before and after interconnection, the effectiveness of this strategy was quantitatively assessed.

At Yugye Reservoir in Gongju, daily water balance simulations over 47 years (1971–2017) revealed that water levels fell to the dead low water level in four years (1994, 1995, 2015, and 2016), resulting in a water supply reliability of 91.5%. When surplus water from Wolgok Reservoir was supplied, reliability increased to 100%. Although Wolgok has only one-fourth the effective storage of Yugye, its watershed area is 1.3 times larger, enabling the redirection of an average surplus discharge of 2338.4 × 10³ m³. This effectively compensated for Yugye’s low watershed ratio. Following interconnection, the minimum annual water level rose to 66.62 EL.m, which is 1.62 m above the dead storage threshold (65.0 EL.m), clearly indicating enhanced supply capacity. The average annual maximum water shortage prior to the interconnection was 15.5 × 10³ m³, whereas no shortage was observed following the interconnection. Considering the distance between reservoirs and the estimated costs of comparable projects in 2024, the total project cost is expected to be approximately 260 million KRW. This reservoir is considered suitable for integrated operation with other reservoirs, as it can alleviate water shortages at a relatively low cost.

In Boryeong, Okgye Reservoir’s reliability improved by 2.2% when supplied with surplus water from Cheongcheon Reservoir, and the average annual maximum water shortage was reduced from 60.3 × 10³ m³ to 37.9 × 10³ m³. The upstream location of Okgye Reservoir relative to Cheongcheon Reservoir enables the reuse of agricultural water, as it can flow back into Cheongcheon Reservoir following interconnection. In terms of project cost, it is estimated to be approximately twice that of the Yugye Reservoir.

Dongbu Reservoir in Seocheon showed a 10% reliability gain when connected to Munsan Reservoir, located just 2.5 km away, demonstrating the benefit of short-range interconnection. The average annual maximum water shortage decreased by more than 2× 10⁶ m³, from 6827.2 × 10³ m³ to 4278.2 × 10³ m³, while the project cost is estimated to be approximately four times higher than that of the Yugye Reservoir. Likewise, Songak Reservoir in Cheonan exhibited a reliability increase of approximately 7%. In contrast, Seobu Reservoir showed no improvement in reliability despite receiving surplus water; however, the minimum annual water level increased by 0.14 m, reaching 14.33 EL.m.

6. Discussion

Overall, this study confirms that supplying surplus water from reservoirs with higher watershed ratios can enhance the reliability of those with lower ratios. Among the five analyzed reservoirs, four exhibited an average increase of approximately 7% in water supply reliability, although performance varied depending on geomorphological and structural characteristics. The interconnection of all five reservoirs resulted in a significant reduction in the maximum annual water shortage, and the potential for addressing shortages through water reuse in the Okgye and Songak Reservoirs was confirmed. In this study, a preliminary cost review based on the connection distance indicated that stable water resources could be secured at a relatively low cost. Future research should extend beyond Chungcheongnam-do, incorporating broader topographic variability and investigating seasonal and situational reservoir operation strategies on a national scale. Furthermore, further research is warranted on how to utilize the surplus storage capacity of reservoirs by interconnecting those distributed throughout the country, as well as on the operational strategies for such reservoir networks under varying temporal and situational conditions.