Mobile QR Code QR CODE : Journal of the Korean Society of Civil Engineers

  1. ์ •ํšŒ์›โ€ค์›๊ด‘๋Œ€ํ•™๊ต ๊ฑด์„คํ™˜๊ฒฝ๊ณตํ•™๊ณผ ์„์‚ฌ๊ณผ์ • (Wonkwang Universityโ€คnhg1221@wku.ac.kr)
  2. ์ข…์‹ ํšŒ์›โ€ค๊ต์‹ ์ €์žโ€ค์›๊ด‘๋Œ€ํ•™๊ต ๊ฑด์„คํ™˜๊ฒฝ๊ณตํ•™๊ณผ ๋ถ€๊ต์ˆ˜ (Corresponding Authorโ€คWonkwang Universityโ€คkim2018@wku.ac.kr)



์ผ์‚ฌ๋Ÿ‰, ํŒจ๋„ ํšจ์œจ, ์ˆ˜์ƒํƒœ์–‘๊ด‘, ๋ฐœ์ „๋Ÿ‰, ๋ชฌํ…Œ์นด๋ฅผ๋กœ ์‹œ๋ฎฌ๋ ˆ์ด์…˜
Insolation, Solar panel efficiency, Floating solar power, Generation, Monte Carlo simulation

1. ์„œ ๋ก 

2020๋…„ ๋Œ€ํ•œ๋ฏผ๊ตญ ์‹ ์žฌ์ƒ์—๋„ˆ์ง€ ๋ณด๊ธ‰ํ†ต๊ณ„์— ์˜ํ•˜๋ฉด, ์‹ ์žฌ์ƒ์—๋„ˆ์ง€์˜ ๋ฐœ์ „๋Ÿ‰์€ 43,062 GWh๋กœ ๊ทธ์ค‘ ํƒœ์–‘๊ด‘์€ 19,298 GWh (44.8 %)์ด๋‹ค(KEA, 2021). ์‹ ์žฌ์ƒ์—๋„ˆ์ง€ ์ „์ฒด ์„ค๋น„์šฉ๋Ÿ‰์€ 25 GW์ด๋ฉฐ, ํƒœ์–‘๊ด‘์ด 17 GW (66.8 %)๋กœ ๊ฐ€์žฅ ๋†’์œผ๋ฉฐ, ๋ฐ”์ด์˜ค๋Š” 3.5 GW (13.6 %), ์ˆ˜๋ ฅ์€ 1.8 GW (7 %), ํ’๋ ฅ์€ 1.6 GW (6 %)์ด๋‹ค(KEA, 2021). Fig. 1์—์„œ์™€ ๊ฐ™์ด ์‹ ์žฌ์ƒ์—๋„ˆ์ง€ ๋ฐœ์ „๋Ÿ‰๊ณผ ์„ค๋น„์šฉ๋Ÿ‰ ๋ชจ๋‘ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์ด ๊ฐ€์žฅ ๋งŽ์€ ์ด์œ ๋Š” ์ง€์ƒ ๋ฐ ์ˆ˜์ƒ์— ์„ค์น˜๊ฐ€ ๊ฐ€๋Šฅํ•˜๊ณ , ์ž‘์€ ๋ฉด์ ์—๋„ ์„ค์น˜๊ฐ€ ๊ฐ€๋Šฅํ•œ ์šฐ๋ฆฌ๋‚˜๋ผ์— ์ ํ•ฉํ•œ ๋ฐœ์ „ ๋ฐฉ์‹์ด๊ธฐ ๋•Œ๋ฌธ์ด๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์‹œ์„ค์„ ๊ฑด์ถ•๋ฌผ๊ณผ ํ† ์ง€์— ์„ค์น˜ํ•˜๋Š” ๊ฒฝ์šฐ ์„ค์น˜ ๋ฉด์ , ๊ตฌ์กฐ ์•ˆ์ •์„ฑ, ์ž„๋Œ€๋น„์šฉ, ๋ฏผ์›๊ณผ ์ธํ—ˆ๊ฐ€ ๊ด€๋ จ ๋ฌธ์ œ ๋“ฑ์ด ๋ฐœ์ƒํ•œ๋‹ค(Joo, 2014). ํ•˜์ง€๋งŒ ์ˆ˜์ƒํƒœ์–‘๊ด‘์€ ์ด๋Ÿฌํ•œ ๋ฌธ์ œ์ ๋“ค์„ ํ•ด๊ฒฐํ•˜๋Š” ๋Œ€์•ˆ์ด ๋  ์ˆ˜ ์žˆ๋‹ค.

์ˆ˜์ƒํƒœ์–‘๊ด‘์€ ํƒœ์–‘๊ด‘ ๋ชจ๋“ˆ์„ ์ˆ˜๋ฉด ์œ„์— ์„ค์น˜ํ•œ ๋ฐœ์ „์‹œ์Šคํ…œ์œผ๋กœ ์œ ํœด ์ˆ˜๋ฉด์— ์„ค์น˜๋˜๊ธฐ ๋•Œ๋ฌธ์— ํ† ์ง€ ๊ด€๋ จ ๋ฌธ์ œ๊ฐ€ ๋ฐœ์ƒํ•˜์ง€ ์•Š๋Š”๋‹ค. ์ˆ˜์ƒํƒœ์–‘๊ด‘์€ ์„ค๋น„์˜ ๊ตฌ์กฐ์— ๋”ฐ๋ผ์„œ ์ผ๋ฐ˜์ ์œผ๋กœ ์„ธ ๊ฐ€์ง€๋กœ ๋ถ„๋ฅ˜ํ•˜๋Š”๋ฐ, ํƒœ์–‘๊ด‘ ๋ชจ๋“ˆ ์ง€์ง€์ฒด์™€ ๋ถ€๋ ฅ์ฒด๊ฐ€ ์ผ์ฒด๋œ ํฐํŠ ํ˜•, ๋ถ„๋ฆฌ๋œ ํ”„๋ ˆ์ž„ํ˜•, ํƒœ์–‘์„ ์ถ”์ ํ•˜์—ฌ ํšŒ์ „ํ•˜๋Š” ํƒœ์–‘ ์ถ”์ ํ˜• ๋“ฑ์ด๋‹ค(Ahn et al., 2021). ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ๋Š” 2022๋…„ 1์›” ์ „๊ตญ์— 183 MW ์‹œ์„ค์šฉ๋Ÿ‰์ด ์„ค์น˜๋˜์–ด ์žˆ์œผ๋ฉฐ, ๊ทธ์ค‘ ์ €์ˆ˜์ง€์— ์„ค์น˜๋œ ๊ฒƒ์€ 104.768 MW์ด๊ณ , ๋Œ์— ์„ค์น˜๋œ ๊ฒƒ์€ 47.167 MW์ด๋‹ค(Kwon et al., 2022b). ์ตœ๊ทผ์— ๋Œ€์šฉ๋Ÿ‰์˜ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „์‹œ์„ค์„ ํ•ด์ƒ์— ์„ค์น˜ํ•˜๋ ค๋Š” ๊ณ„ํš์ด ๊ตญ๊ฐ€ ์ฃผ๋„๋กœ ์ง„ํ–‰๋˜๊ณ  ์žˆ๋‹ค. ์ „๋ผ๋ถ๋„ ๊ตฐ์‚ฐ, ๊น€์ œ, ๋ถ€์•ˆ์ง€์—ญ์˜ ์ƒˆ๋งŒ๊ธˆ๊ฐœ๋ฐœ์ง€์—ญ์˜ ํ•ด์ƒ์— 2.1 GW์˜ ์‹œ์„ค์šฉ๋Ÿ‰์œผ๋กœ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „๋‹จ์ง€ ์กฐ์„ฑ ๊ณ„ํš์— ์žˆ๋‹ค(SDIA, 2022).

ํŠนํžˆ, ๊ตฐ์‚ฐ ์ˆ˜์ƒํƒœ์–‘๊ด‘์€ ์ „๋ถ ๊ตฐ์‚ฐ์‹œ ๋น„์‘๋„ ๊ตญ๊ฐ€์‚ฐ๋‹จ์— ์œ„์น˜ํ•˜๋ฉฐ, ๋‹น์‹œ ์ตœ๋Œ€์‹œ์„ค์šฉ๋Ÿ‰์ธ 18.7 MW์œผ๋กœ ์„ค์น˜ ์œ ์ˆ˜์ง€ ๋ฉด์ ์€ 221,400 m2์œผ๋กœ ์ด์‚ฌ์—…๋น„ 431์–ต์›์ด ํˆฌ์ž…๋œ ์‚ฌ์—…์ด๋‹ค(KOEN, 2021). ๋ฐœ์ „ ์šด์˜๊ธฐ๊ฐ„์€ ์ค€๊ณต ํ›„ 20๋…„๊ฐ„์ด๋ฉฐ, ๊ตญ์‚ฐ ํƒœ์–‘๊ด‘๋ชจ๋“ˆ 51,912์žฅ, ์ธ๋ฒ„ํ„ฐ 1 MW๊ธ‰ 19๋Œ€, ์ˆ˜์ƒํƒœ์–‘๊ด‘์„ ์ง€์ง€ํ•˜๊ณ  ์žˆ๋Š” PE๋ถ€๋ ฅ์ฒด 25,925์žฅ ๋“ฑ ์‹œ์„ค์„ ๊ฐ–์ถ”๊ณ  ์žˆ๋‹ค(KOEN, 2021). ์‹œ์„ค๋“ค์€ ํƒ€ ์‚ฌ์—…๋ณด๋‹ค ๊ตญ์‚ฐํ™”์œจ์ด ๋†’์œผ๋ฉฐ, ์‚ฌ์—…์ข…๋ฃŒ ํ›„์— 100 % ์žฌํ™œ์šฉ์ด ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๊ฒƒ์ด ํฐ ํŠน์ง•์ด๋‹ค(KOEN, 2021). 2018๋…„ 9์›” ์ฒซ ์ƒ์—…์šด์ „๋˜์—ˆ์œผ๋ฉฐ, ๊ฑด์„ค ์ „ ์˜ˆ์ƒํ•œ ์—ฐ๊ฐ„ ๋ฐœ์ „๋Ÿ‰์€ 25,322 MWh์ด๋‹ค(์›”ํ‰๊ท  ์•ฝ 2.1 GWh)์ด๋‹ค(KOEN, 2021).

์ˆ˜์ƒํƒœ์–‘๊ด‘์„ ํฌํ•จํ•œ ํƒœ์–‘๊ด‘์˜ ๊ฒฝ์šฐ, ๋ฐœ์ „๋Ÿ‰์€ ๊ธฐ์ƒ ๋ฐ์ดํ„ฐ์— ๋ฐ€์ ‘ํ•œ ์˜ํ–ฅ์„ ๋ฐ›๊ณ  ์žˆ์œผ๋ฉฐ, ๊ธฐ์ƒ์กฐ๊ฑด์— ๋”ฐ๋ผ ๋ณ€๋™์„ฑ์ด ๋งค์šฐ ํฌ๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์€ ์•ˆ์ •์ ์œผ๋กœ ์ „๋ ฅ์„ ๊ณต๊ธ‰ํ•˜๊ณ , ๊ฒฝ์ œ์ ์ธ ์ˆ˜์ต์„ ์–ป๊ธฐ ์œ„ํ•ด ๋งค์šฐ ์ค‘์š”ํ•˜๋‹ค. ๋ฐœ์ „์†Œ์˜ ์šด์˜์„ ์œ„ํ•œ ๊ณ„ํš๊ณผ ๋ถˆํ™•์‹ค์„ฑ์— ๋Œ€ํ•œ ์˜ˆ์ธก์„ ์œ„ํ•ด์„œ๋Š” ๋ฐœ์ „๋Ÿ‰์˜ ์ •ํ™•ํ•œ ์˜ˆ์ธก์ด ํ•„์š”ํ•˜๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์‚ฐ์ •์— ํ•„์š”ํ•œ ์š”์†Œ๋“ค์€ ์ผ์‚ฌ๋Ÿ‰, ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ๊ณผ ํƒœ์–‘๊ด‘ ํŒจ๋„ ์„ค์น˜ ๋ฉด์  ๋“ฑ์ด๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” ๊ธฐ์ƒ์ฒญ ๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์—ฌ ๊ตฐ์‚ฐ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „์‹œ์„ค์˜ ๊ฑด์„ค ๋‹น์‹œ ์ถ”์ • ๋ฐœ์ „๋Ÿ‰๊ณผ ๊ณผ๊ฑฐ ์ผ์‚ฌ๋Ÿ‰์˜ ๋ณ€๋™์„ ๊ณ ๋ คํ•œ ๋ฏธ๋ž˜ ๋ฐœ์ „๋Ÿ‰์„ ํ™•๋ฅ ์ ์œผ๋กœ ๋ถ„์„ํ•˜์—ฌ ๋น„๊ตํ•˜๊ณ ์ž ํ•œ๋‹ค. ์ด๋Ÿฌํ•œ ๋น„๊ต๋ถ„์„์„ ํ†ตํ•˜์—ฌ ํƒœ์–‘๊ด‘๋ฐœ์ „ ์‚ฌ์—…์„ ์ˆ˜ํ–‰ํ•˜๋Š”๋ฐ ์žˆ์–ด์„œ ์ค‘์š”ํ•œ ์š”์†Œ์ธ ์„ค์น˜ ์ „ ๋ฐœ์ „๋Ÿ‰ ์ถ”์ •์— ๊ฐœ์„  ๋ฐฉํ–ฅ์„ ์ œ์‹œํ•˜๊ณ ์ž ํ•œ๋‹ค.

Fig. 1. Renewable Energy in South Korea (KEA, 2021)
../../Resources/KSCE/Ksce.2023.43.2.0249/fig1.png

2. ์„ ํ–‰์—ฐ๊ตฌ ๋ถ„์„

๊ธฐํ›„์ธ์ž์— ์˜ํ–ฅ์„ ๋งŽ์ด ๋ฐ›๋Š” ํƒœ์–‘๊ด‘ ๋ฐœ์ „์‹œ์„ค์˜ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•œ ์—ฐ๊ตฌ๋Š” ๋งŽ์€ ์—ฐ๊ตฌ์ž๋“ค์— ์˜ํ•ด ์‹œ๋„๋˜์—ˆ๋‹ค. ๊ทธ์ค‘์—์„œ ์ตœ๊ทผ์˜ ์„ ํ–‰ ์—ฐ๊ตฌ๋“ค์„ ๋ถ„์„ํ•˜์˜€๋‹ค(Table 1). Lee et al.(2022)์€ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์„ ์œ„ํ•ด ์ผ์‚ฌ๋Ÿ‰, ๋ชจ๋“ˆ ์˜จ๋„ ๋ฐœ์ „๋Ÿ‰ ๋“ฑ์„ ๋ถ„์„ํ•˜์˜€์œผ๋ฉฐ, ์˜ˆ์ธก ๋ชจ๋ธ์„ ๋งŒ๋“ค๊ธฐ ์œ„ํ•ด ๊ธฐ์˜จ, ์Šต๋„ ๋“ฑ์˜ ๊ธฐ์ƒ์ฒญ ๋‹จ๊ธฐ์˜ˆ๋ณด ์ •๋ณด๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ํŒŒ์ด์ฌ ํ”„๋กœ๊ทธ๋žจ์„ ์‚ฌ์šฉํ•œ ํ†ต๊ณ„์  ๋ชจํ˜• ๊ธฐ๋ฐ˜ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก ๋ชจ๋ธ๊ณผ ์ธ๊ณต์ง€๋Šฅ ๋ชจํ˜• ๊ธฐ๋ฐ˜ ์˜ˆ์ธก ๋ชจ๋ธ์„ ๊ตฌํ˜„ํ•˜๊ณ , ๊ฐ ๋ชจ๋ธ์˜ ๊ฒฐ๊ณผ๋ฅผ ๋น„๊ต, ๋ถ„์„ํ•˜์˜€๋‹ค. Gastli and Charabi(2010)์€ Oman์˜ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ์ „๋ง์„ ์—ฐ๊ตฌํ•˜๊ธฐ ์œ„ํ•ด, GIS๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰ ์ง€๋„๋ฅผ ์ƒ์„ฑํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์‹œํ•˜์˜€๋‹ค. Kwon et al.(2022a)์€ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด ํ•ฉ์ฒœ๋Œ 500 kW ์‹œ์„ค์šฉ๋Ÿ‰์˜ ๋ฐœ์ „๋Ÿ‰๊ณผ ์ผ์‚ฌ๋Ÿ‰, ์˜จ๋„, ํ’์†, ๊ฐ•์šฐ ๋“ฑ์˜ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜๊ณ , ์ˆœํ™˜์‹ ๊ฒฝ๋ง์„ ์ด์šฉํ•˜์—ฌ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜์˜€๋‹ค. Lee and Song(2022)์€ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์„ ์œ„ํ•ด 1์‹œ๊ฐ„ ๋‹จ์œ„ ๋ฐœ์ „๋Ÿ‰๊ณผ ๊ธฐ์ƒ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ๋”ฅ๋Ÿฌ๋‹๊ณผ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ธฐ๋ฒ•์„ ์ ์šฉํ•œ ์˜ˆ์ธก ๋ชจ๋ธ์„ ์„ค๊ณ„ํ•˜๊ณ , ๋น„๊ต ๋ถ„์„ํ•˜์˜€๋‹ค. Kim et al.(2017)์€ ์ผ๋ณ„ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ์˜ˆ์ธก ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. ์ œ์•ˆ๋œ ๋ชจ๋ธ์€ ๋ณด์ • ๊ณ„์ˆ˜๋ฅผ ๊ฐ–๋Š” ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ์ถœ๋ ฅ ๊ณต์‹์„ ๊ธฐ๋ฐ˜์œผ๋กœ ํ•˜๋ฉฐ, ๊ธฐ์ƒ ์˜ˆ๋ณด๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ ๋ณด์ • ๊ณ„์ˆ˜๋ฅผ ์ถ”์ •ํ•˜์˜€๋‹ค. 27.34 kW ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ์˜ ์‹ค์ฆ์‹คํ—˜์„ ํ†ตํ•ด,

Table 1. Literature Review of Methods and Data for Solar Power Generation

Authors

Case study

Data type

Method

Lee et al.(2022)

30 kW facility installed in building parking lot, South Korea

Data for learning predictive models: insolation, module temperature, power generation

Data for implementing a forecast model: temperature, humidity, and short-term forecast from the KMA

CNN-LSTM (Convolutional Long Short-Term Memory Neural Networks) model,

moving average,

ARIMA (Autoregressive Integrated Moving Average) model, SARIMA (Seasonal ARIMA) model

Gastli and Charabi(2010)

Oman

Power generation potential, solar radiation received per unit horizontal area, total area, available area, solar system convert efficiency

GIS tools

Kwon et al.(2022a)

500 kW floating solar energy installed in Hapcheon Dam, South Korea

Power generation, weather data

RNN (Artifical Neural Network)

Lee and Song(2022)

South Korea

Power generation, date and time, temperature, wind speed, wind direction, humidity, cloudiness

ANN (Artificial Neural Network), LR (Linear Regression), RF (Random Forest), XGboost, SVM (Support Vector Machine), decision tree

Kim et al.(2017)

Solar power installed at GERI (Gumi Electronics & Information Technology Research Institute), South Korea

Daily weather forecast, the historical data of power generation

New prediction model based on deep learning

Shin et al.(2017)

South Korea

Wind direction, wind speed, humidity, cloudiness

DNN (Deep Neural Network)

Serttas et al.(2018)

Turkey

Power generation

Mycielski - Markov

Song et al.(2014)

South Korea

Weather data for solar insolation prediction

Solar power prediction model

Lee and Kim(2016)

South Korea

Power generation data, weather data collected from the KMA, solar altitude

SVR (Support Vector Regression)

Jeong and Chae(2018)

Solar power in a small building located in Gyeonggi-do, South Korea

Outdoor air temperature, relative humidity, wind direction, wind speed, sky code, rainfall probability

RF (Random Forest), ANN (Artificial Neural Network), SVM (Support Vector Machine)

Kim(2019)

Busan, South Korea

Temperature, wind speed, humidity, Installation angle of panel, tuning time, solar radiation, daylight,

Linear regression, SVR (Support Vector Regression), K-NN, MLP Regression, RFR (Random Forest Regression), AdaBoost, Gradient boosting

Lee and Lee(2016)

South Korea

Climate information, power generation

SVM (support vector machine), neural network, deep learning, algorithm evaluation as RMSE (Root Mean Square Error) value

Kim et al.(2021)

South Korea

Forecast date, rainfall probability, humidity, three-hour temperature, wind speed, forecast time, six-hour rainfall, wind direction, applicable place

Machine learning ensemble forecast-random forest algorithm

Song et al.(2022)

Solar power in Gwangju and Daegu, South Korea

Maximum of hourly climate data (temperature, wind speed, humidity, cloudiness), solar data

TransGRU with Transformer encoder and Gate Circulation Unit

Kang et al.(2022)

South Korea

Insolation, cloudiness, temperature, rainfall, wind speed, sunshine, humidity, status of solar power from EWP (Korea East-West Power)

MLR (Multi Linear Regression) model

Lee and Ji(2015)

Solar power system in Cheongju, South Korea

Power generation, insolation, sunshine duration, cloudiness

ELM (Extreme Learning Machine)

์ œ์•ˆ๋œ ๋ชจ๋ธ์ด ๊ธฐ์กด ๋ชจ๋ธ๋ณด๋‹ค ๋” ์ž˜ ์ž‘๋™ํ•จ์„ ํ™•์ธํ•˜์˜€๋‹ค. Shin et al.(2017)์€ ์ผ์กฐ์‹œ๊ฐ„ ๋ฐ ์ผ์‚ฌ๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด ๊ธฐ์ƒ์ฒญ ๋ฐ์ดํ„ฐ์ธ ๊ฐ•์ˆ˜๋Ÿ‰, ํ’ํ–ฅ, ํ’์†, ์Šต๋„, ์šด๋Ÿ‰ ๋“ฑ์„ ์ด์šฉํ•˜๊ณ , ์˜ˆ์ธก๋œ ์ผ์กฐ์‹œ๊ฐ„ ๋ฐ ์ผ์‚ฌ๋Ÿ‰์„ ์‚ฌ์šฉํ•˜์—ฌ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์‹ค์‹œ๊ฐ„์œผ๋กœ ์˜ˆ์ธกํ•˜๋Š” ๋”ฅ๋Ÿฌ๋‹ ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. Serttas et al.(2018)์€ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ๋ชจ๋“ˆ์˜ ์ถœ๋ ฅ์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด ํ•˜์ด๋ธŒ๋ฆฌ๋“œ ์˜ˆ์ธก ๋ชจ๋ธ์ธ Mycielski-Markov๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์‹œ๊ฐ„๋‹น ์ธก์ •ํ•˜์—ฌ ์ˆ˜์ง‘ํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์˜€๋‹ค. Song et al.(2014)์€ ์‹ค์‹œ๊ฐ„ ๊ธฐ์ƒ ๋ฐ์ดํ„ฐ ์˜ˆ์ธก๊ฐ’์„ ์ด์šฉํ•˜์—ฌ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ์˜ˆ์ธก๊ฐ’์„ ์‚ฐ์ถœํ•˜๋Š” ๋ฐœ์ „ ์˜ˆ์ธก ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด ํƒœ์–‘๊ด‘ ๋ฐœ์ „๊ธฐ๋ณ„ ๋ชจ๋“ˆ์˜ ํŠน์„ฑ ๋ฐ ์˜จ๋„๋ฅผ ๊ณ ๋ คํ•œ ๋ณด์ •๊ณ„์ˆ˜๋ฅผ ์ž…๋ ฅํ•˜๊ณ  ์˜ˆ์ธก์˜์—ญ์˜ ์œ„์น˜๊ฒฝ์‚ฌ๊ฐ์„ ๋ถ„์„ํ•˜์˜€์œผ๋ฉฐ, ๋ฐœ์ „์˜ˆ์ธก๊ณ„์‚ฐ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์—ฐ๊ตฌํ•˜์˜€๋‹ค. Lee and Kim(2016)์€ ๊ธฐ๊ณ„ํ•™์Šต ๋ฐฉ๋ฒ•์ธ ์„œํฌํŠธ ๋ฒกํ„ฐ ํšŒ๊ท€(Support Vector Regression)๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰ ์˜ˆ์ธก ๋ชจ๋ธ์„ ๋งŒ๋“ค๊ณ , ๋ฐœ์ „๋Ÿ‰, ๊ธฐ์ƒ์‹ค์ธก, ๊ธฐ์ƒ์˜ˆ๋ณด ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๋Š” ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. Jeong and Chae(2018)์€ ์ „๋ ฅ ์ถœ๋ ฅ์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•˜์—ฌ ๊ธฐ๊ณ„ํ•™์Šต ๋ชจ๋ธ ์„ ํƒ๋ฐฉ๋ฒ•์„ ์—ฐ๊ตฌํ•˜์˜€์œผ๋ฉฐ, ๋จธ์‹ ๋Ÿฌ๋‹ ์•Œ๊ณ ๋ฆฌ์ฆ˜์ธ RF (Random Forest), ANN (Artificial Neural Network), SVM (Support Vector Machine)์œผ๋กœ ์ ‘๊ทผ๋ฐฉ์‹์„ ํ…Œ์ŠคํŠธ ํ•œ ํ›„ ๊ธฐ์ƒ์ฒญ์˜ ๊ธฐ์กด ๊ธฐ์ƒ์˜ˆ๋ณด ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์˜ˆ์ธก์„ ํ–ˆ๋‹ค. Kim(2019)์€ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์„ ์ž๋™ํ™”ํ•˜๋Š” ๊ธฐ์ˆ ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. ๊ณผ๊ฑฐ ๋ฐœ์ „๋Ÿ‰๊ณผ ๊ธฐ์ƒ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๋จธ์‹ ๋Ÿฌ๋‹ ๋ชจ๋ธ์„ ํ›ˆ๋ จ์„ ํ†ตํ•ด ์ตœ์ ์˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ฐพ๊ณ , ๊ธฐ์ƒ ์˜ˆ๋ณด๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ๋ฏธ๋ž˜ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜์˜€๋‹ค. ๊ธฐ์ƒ ๋ณ€์ˆ˜ ์ ์šฉ์— ๋”ฐ๋ผ ๋ชจ๋ธ ์„ฑ๋Šฅ์˜ ํ–ฅ์ƒ์„ ์œ„ํ•ด ๋ฐœ์ „ ์˜ˆ์ธก์— ์–ด๋–ค ๊ธฐ์ƒ ๋ณ€์ˆ˜๊ฐ€ ๊ฐ€์žฅ ํฐ ์˜ํ–ฅ์„ ์ฃผ๋Š” ์ง€๋ฅผ ๋น„๊ต, ๋ถ„์„ํ•˜์˜€๋‹ค. Lee and Lee(2016)์€ ๊ธฐํ›„ ๋ฐ์ดํ„ฐ์™€ ๋ฉ”์‚ฌ์ถ”์„ธ์ธ ์ฃผ์˜ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ์™€ Concord Cape ๋ฐœ์ „์†Œ์˜ ๋ฐœ์ „๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ SVM (Support Vector Machine), ์‹ ๊ฒฝ๋ง, ๋”ฅ๋Ÿฌ๋‹์„ ์ด์šฉํ•ด ์‹คํ—˜ํ•˜์˜€์œผ๋ฉฐ, RMSE (Root Mean Square Error)์œผ๋กœ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ํ‰๊ฐ€ํ•œ ๋’ค ์ตœ์ ์˜ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์„ ์ •ํ•˜์—ฌ ๊ตญ๋‚ด ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ์— ์ ์šฉํ•˜์˜€๋‹ค. Kim et al.(2021)์€ ๋ฐœ์ „๋Ÿ‰ ์žฅ๊ธฐ ์˜ˆ์ธก์„ ์œ„ํ•ด ๊ธฐ๊ณ„ํ•™์Šต ๊ธฐ๋ฒ•์ธ ๋žœ๋คํฌ๋ ˆ์ŠคํŠธ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ์˜ค์ฐจ๋ฅผ ์ค„์ด๊ณ  ์ˆ˜์ต์˜ ์ตœ๋Œ€๋ฅผ ์œ„ํ•ด ๊ธฐ์ƒ ์˜ˆ๋ณด ๋ฐ์ดํ„ฐ์™€ ์‹ค์ œ ๋‚ ์”จ ์ž๋ฃŒ๋ฅผ ์ด์šฉํ•˜์—ฌ ์—ฐ๊ตฌ๋ฅผ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. Song et al.(2022)์€ ๊ธฐ์ƒ๊ด€์ธก์†Œ๋ฅผ ํ†ตํ•ด ๋“ค๋กœ๋„ค ์‚ผ๊ฐ๋ถ„ํ• ์„ ์ด์šฉํ•˜์—ฌ ๊ตญ๋‚ด ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๋Š” ์ƒˆ๋กœ์šด ๊ธฐ์ˆ ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. ๊ธฐ์ƒ ์˜ˆ๋ณด ์˜ค์ฐจ์— ๊ฐ•ํ•˜๊ณ , ์˜ˆ์ธก ์„ฑ๋Šฅ์ด ๊ฐœ์„ ๋œ Transformer encoder์™€ Gate Circulation Unit (GRU; ๊ฒŒ์ดํŠธ ์ˆœํ™˜ ์œ ๋‹›)์„ ์ด์šฉํ•˜์—ฌ ์ƒˆ๋กœ์šด TransGRU ๋ชจ๋ธ์„ ๊ตฌํ˜„ํ•˜๊ณ , ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก ์—ฐ๊ตฌ๋ฅผ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. Kang et al.(2022)์€ ๋จธ์‹ ๋Ÿฌ๋‹ ๋ชจ๋ธ์ธ ๋‹ค์ค‘์„ ํ˜•ํšŒ๊ท€ ๋ชจ๋ธ(Multi Linear Regression)์„ ์ด์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ณ , ์˜ˆ์ธกํ•œ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก ๊ธฐ๋ฒ•์„ ๊ฐœ๋ฐœํ•˜๊ณ , DC ๋‚˜๋…ธ๊ทธ๋ฆฌ๋“œ ํ™˜๊ฒฝ์—์„œ ๊ฐ„ํ—์ ์ธ ์ถœ๋ ฅ์„ ๊ฐ–๋Š” ์‹ ์žฌ์ƒ ์—๋„ˆ์ง€์›์˜ ์ ํ•ฉ์„ฑ์„ ๋ถ„์„ํ•˜์˜€๋‹ค. Lee and Ji(2015)์€ ํ˜„์žฌ ์šด์˜ ์ค‘์ธ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ์—์„œ 1๋…„๊ฐ„ ์ธก์ •ํ•œ ๋ฐœ์ „๋Ÿ‰ ๋ฐ์ดํ„ฐ์™€ ๊ธฐ์ƒ์ •๋ณด์˜ ์ƒ๊ด€๊ด€๊ณ„๋ฅผ ๋ถ„์„ํ•˜์—ฌ ๋†’์€ ์ƒ๊ด€๊ด€๊ณ„ ์ž…๋ ฅ์š”์ธ์„ ๋„์ถœํ•˜์˜€์œผ๋ฉฐ, ์„ ์ •๋œ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ตœ์ƒํ•™์Šต๊ธฐ๊ณ„(ELM: Extreme Learning Machine)๋ฅผ ์ด์šฉํ•œ ์ผ์ผ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์ „๋ ฅ ์˜ˆ์ธก ๋ชจ๋ธ์„ ์ œ์‹œํ•˜์˜€๋‹ค.

Table 2. Literature Review of MCS for Solar Power Generation

Authors

Country

Case study

Data type

Kim et al.(2008)

South Korea

Farm land in Seosan-gun, Chungcheongnam-do

Probability distribution of solar radiation

Liu et al.(2018)

China

Virtual case

Power generation

Ko and Kim(2020)

South Korea

Virtual case

Solar power generation assumption

Jeon et al.(2019)

South Korea

Solar and wind power plants

Insolation, wind speed, temperature

Ryu et al.(2020)

South Korea

Simulation system of Gangseo-gu in Seoul

Historical climate data

Liew and Lee(2020)

South Korea

Assume hybrid plant installation in major region

Irradiance data

Kim and Ryu(2020)

South Korea

Virtual case

Energy supply and demand (100 kW-200 kW random data)

Lee and Kim(2020)

South Korea

Virtual case

Generator and load data from WT modeling as input data

ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•œ ์„ ํ–‰ ์—ฐ๊ตฌ ์ค‘์—์„œ ๋Œ€ํ‘œ์ ์ธ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ถ„์„๋ฐฉ๋ฒ•์ธ ๋ชฌํ…Œ์นด๋ฅผ๋กœ ์‹œ๋ฎฌ๋ ˆ์ด์…˜(Monte Carlo Simulation: MCS) ๊ธฐ๋ฒ•์„ ํ™œ์šฉํ•œ ์—ฐ๊ตฌ๋“ค์— ๋Œ€ํ•ด์„œ ๋ฐฉ๋ฒ•๋ก ๊ณผ ์ž๋ฃŒ๋“ค์— ๋Œ€ํ•ด ๊ฒ€ํ† ํ•˜๊ณ  ๋ถ„์„ํ•˜์˜€๋‹ค(Table 2). Kim et al.(2008)์€ ํŠน์ • ํƒœ์–‘๊ด‘ ์‹œ์Šคํ…œ์„ ๊ธฐ์ค€์œผ๋กœ ์ผ๋ณ„ ์ผ์‚ฌ๋Ÿ‰์„ ๋ฐœ์ „๋Ÿ‰์œผ๋กœ ํ™˜์‚ฐํ•˜๊ณ , ์ถฉ์ฒญ๋‚จ๋„ ์„œ์‚ฐ๊ตฐ ๋†์ดŒ์ง€์—ญ์˜ ์›”๋ณ„ ๋ฐœ์ „๋Ÿ‰์˜ ์ ์ • ํ™•๋ฅ ๋ถ„ํฌ๋ฅผ ์ •์˜ํ•˜๊ธฐ ์œ„ํ•ด MCS๋ฅผ ์ด์šฉํ•˜์˜€๋‹ค. ํ•ด๋‹น ์—ฐ๊ตฌ์—์„œ๋Š” ๋†์ดŒ๋งˆ์„ ํƒœ์–‘๊ด‘๋ฐœ์ „์‹œ์Šคํ…œ ๊ณ„ํš์„ ์ˆ˜๋ฆฝํ•  ๋•Œ ๋ฐœ์ „๋Ÿ‰์˜ ํ™•๋ฅ ๋ถ„ํฌ๋ฅผ ๊ณ ๋ คํ•˜๋Š” ๊ฒƒ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. Liu et al.(2018)์€ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ ์ „๋ ฅ ์‹œ์Šคํ…œ์˜ ์‹ ๋ขฐ์„ฑ์„ ํ‰๊ฐ€ํ•˜๊ธฐ ์œ„ํ•˜์—ฌ, ์ˆœ์ฐจ MCS (PMCS: Pseudo-sequential Monte Carlo simulation)์™€ ISSR (Intelligent State Space Reduction)์„ ๊ฒฐํ•ฉํ•œ ์ƒˆ๋กœ์šด ๋ฐฉ๋ฒ•๋ก ์„ ์ œ์•ˆํ•˜์˜€๋‹ค. Ko and Kim(2020)์€ K-Means์™€ K-Medoid ๊ตฐ์ง‘๋ถ„์„์„ ์ด์šฉํ•˜์—ฌ ์‹œ๋‚˜๋ฆฌ์˜ค๋ฅผ ์ถ•์•ฝ๋ฐฉ๋ฒ•์— ๋Œ€ํ•ด ์ œ์•ˆํ•˜๊ธฐ ์œ„ํ•ด, MCS๋กœ 1,000๊ฐœ์˜ ๋‚œ์ˆ˜๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ƒ์„ฑ๋œ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ๋ถˆํ™•์‹ค์„ฑ ์‹œ๋‚˜๋ฆฌ์˜ค๋“ค์„ 10๊ฐœ์˜ ์‹œ๋‚˜๋ฆฌ์˜ค๋กœ ์ถ•์•ฝํ•˜๊ณ , ๋‹ค์‹œ 2๊ฐœ์˜ ํด๋Ÿฌ์Šคํ„ฐ ๋ถ„์„ ๋ฐฉ๋ฒ•์„ ์‚ฌ์šฉํ•˜์—ฌ ๋„์ถœ๋œ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ์‹œ๋‚˜๋ฆฌ์˜ค์™€ ํ™•๋ฅ ์„ ๋ถ„์„ํ•˜์˜€๋‹ค. Jeon et al.(2019)์€ ์ „๋ ฅ ์‹œ์Šคํ…œ์˜ ์•ˆ์ •์ ์ธ ์šด์˜์„ ์œ„ํ•ด ์ค‘์š”ํ•œ ์š”์†Œ์ธ ์ˆœ๋ถ€ํ•˜์˜ ๋ถˆํ™•์‹ค์„ฑ๊ณผ ๋ณ€๋™์„ฑ์„ ๋ถ„์„ํ•˜์˜€๋‹ค. ์ˆœ๋ถ€ํ•˜๋ฅผ ์ถ”์ •ํ•˜๊ธฐ ์œ„ํ•ด ๋‘ ๊ฐ€์ง€ ๋ฐฉ๋ฒ•์œผ๋กœ 2๋‹จ๊ณ„ ARMAX ๋ชจ๋ธ๊ณผ MCS๋ฅผ ์ ์šฉํ•˜๊ณ , ๋ถ„์„ํ•˜์—ฌ 2030๋…„ ์ˆœ์ˆ˜์š” ํ”„๋กœํŒŒ์ผ์„ ๋„์ถœํ•˜์˜€๋‹ค. Ryu et al.(2020)์€ ์‹ ์žฌ์ƒ์—๋„ˆ์ง€์›์˜ ์ถœ๋ ฅ ์˜ˆ์ธก์— ์ธ๊ณต์‹ ๊ฒฝ๋ง ์ค‘ LSTM (Long Short-term Memory)๊ธฐ์ˆ ์„ ์‚ฌ์šฉํ•˜์˜€๋‹ค. LSTM๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ณ , MCS๋กœ ์ƒ์„ฑํ•œ 1,000๊ฐœ์˜ ์‹œ๋‚˜๋ฆฌ์˜ค๋ฅผ K-Means ํด๋Ÿฌ์Šคํ„ฐ๋ง ๊ธฐ๋ฒ•์œผ๋กœ ๊ตฐ์ง‘ ๊ณผ์ •์„ ํ†ตํ•ด 10๊ฐœ์˜ ๊ตฐ์ง‘ ์‹œ๋‚˜๋ฆฌ์˜ค๋ฅผ ์ƒ์„ฑํ•˜์˜€๋‹ค. ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์— ๋”ฐ๋ฅธ ์‹œ์Šคํ…œ์˜ ์œ„ํ—˜์„ฑ ํ‰๊ฐ€๋ฅผ ์ง„ํ–‰ํ•˜์˜€๋‹ค. Liew and Lee(2020)์€ ๋ชฌํ…Œ์นด๋ฅผ๋กœ ๊ด‘ํ•™ ๋ชจ๋ธ๊ณผ ๋ฏธ๊ตญ Solar Electric Generating Station (SEGS) VI ๋ฐœ์ „์†Œ์˜ ์ฆ๊ธฐ ๋™๋ ฅ ์‚ฌ์ดํด ๋ถ„์„ ๋ชจ๋ธ์„ ์ ์šฉํ•˜์—ฌ ํ•˜์ด๋ธŒ๋ฆฌ๋“œ ๊ธฐ์ˆ ์„ ํ‰๊ฐ€ํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์‹œํ•˜์˜€๋‹ค. ๊ตญ๋‚ด ์ฃผ์š”์ง€์—ญ ๊ธฐ์ƒ๋ฐ์ดํ„ฐ๋ฅผ ์ ์šฉํ•˜์—ฌ ์ œ์•ˆ๋œ ํ•˜์ด๋ธŒ๋ฆฌ๋“œ ๊ธฐ์ˆ ์˜ ๋ฐœ์ „๋Ÿ‰๊ณผ ๋ฐ ํšจ์œจ์„ ์‚ฐ์ถœํ•˜๊ณ , ๋™์ผํ•œ ๊ทœ๋ชจ์˜ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ ํ”Œ๋žœํŠธ์™€ ๋น„๊ตํ•˜์—ฌ ๊ตญ๋‚ด ์ ์šฉ ๊ฐ€๋Šฅ์„ฑ์„ ์กฐ์‚ฌํ•˜์˜€๋‹ค. Kim and Ryu(2020)์€ ์—๋„ˆ์ง€ ์ €์žฅ ์‹œ์Šคํ…œ์—์„œ ์•ˆ์ •์  ์šด์˜๊ณผ ํ•„์š”ํ•œ ์šฉ๋Ÿ‰์„ ์‚ฐ์ •ํ•˜๊ธฐ ์œ„ํ•˜์—ฌ, MCS ๊ธฐ๋ฒ•์„ ์‚ฌ์šฉํ•˜์—ฌ ์ˆ˜๊ธ‰์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ๋ฐ˜์˜ํ•˜์—ฌ ๋ถ„์„ํ•˜์˜€๋‹ค. Lee and Kim(2020)์€ ์—๋„ˆ์ง€์ €์žฅ์žฅ์น˜ ์šด์˜์˜ ์—๋„ˆ์ง€ ์ˆ˜๊ธ‰์˜ ๊ณ ์œ ํ•œ ๋ณ€ํ™”๋ฅผ ํŒŒ์•…ํ•˜๊ณ , ์ตœ์ ํ™”๋ฅผ ์ง„ํ–‰ํ•˜๊ธฐ ์œ„ํ•ด MCS๋ฅผ ํ†ตํ•ด ์—๋„ˆ์ง€์ €์žฅ์žฅ์น˜์˜ ํ™•๋ฅ ์  ์šฉ๋Ÿ‰์„ ๊ณ„์‚ฐํ•˜์˜€๋‹ค. ์„ค์ •๋œ ๋ชจ๋ธ๋ง์— ๋”ฐ๋ผ ํƒœ์–‘๊ด‘๋ฐœ์ „ ๋ฐ ๋ถ€ํ•˜๋ฅผ ์ถœ๋ ฅํ•˜๊ณ , MCS๋ฅผ ํ†ตํ•ด ์ƒ์„ฑ๋œ ์‹œ๋‚˜๋ฆฌ์˜ค์˜ ๋ฌด์ž‘์œ„ ์ƒ˜ํ”Œ๋ง์„ ๊ธฐ๋ฐ˜์œผ๋กœ ๋ฌธ์ œ์— ์ ์šฉํ•˜์—ฌ ๊ฒฐ๊ณผ๊ฐ’์˜ ์˜ค๋ฅ˜๋ฅผ ์ตœ์†Œํ™”ํ•˜์˜€๋‹ค.

ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด ๊ธฐํ›„์ธ์ž๋ฅผ ํ™œ์šฉํ•œ ์—ฐ๊ตฌ๋“ค์ด ๋งŽ์•˜๋‹ค. ๋Œ€๋ถ€๋ถ„์˜ ์—ฐ๊ตฌ๋“ค์ด ์‹œ๋ฎฌ๋ ˆ์ด์…˜์œผ๋กœ ์ง„ํ–‰์ด ๋˜์—ˆ์œผ๋ฉฐ, ์ตœ๊ทผ์— ๋”ฅ๋Ÿฌ๋‹, ๋จธ์‹ ๋Ÿฌ๋‹ ๋“ฑ์˜ ์ธ๊ณต์ง€๋Šฅ์„ ํ™œ์šฉํ•œ ๋ฐœ์ „๋Ÿ‰ ๊ด€๋ จ ์—ฐ๊ตฌ๋“ค์ด ์ ์ง€ ์•Š๋‹ค. ์ด๋“ค ์—ฐ๊ตฌ ์—ญ์‹œ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ฐฉ๋ฒ•์„ ํ™œ์šฉํ•˜์—ฌ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ณ  ์žˆ๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” ๊ทธ๋™์•ˆ ์—ฌ๋Ÿฌ ์—ฐ๊ตฌ๋…ผ๋ฌธ์—์„œ ๋งŽ์€ ๊ฒ€์ฆ์ด ๋œ ๋Œ€ํ‘œ์ ์ธ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ถ„์„ ๋ฐฉ๋ฒ•์ธ MCS ๊ธฐ๋ฒ•์„ ์‚ฌ์šฉํ•˜๋ ค๊ณ  ํ•œ๋‹ค. MCS๋Š” ๊ธฐํ›„๋ณ€ํ™”๋กœ ์ธํ•œ ์ผ์‚ฌ๋Ÿ‰ ๋ณ€๋™์˜ ๊ฒฐ๊ณผ๋ฅผ ์ •ํ™•ํ•˜๊ฒŒ ์˜ˆ์ธกํ•˜๊ธฐ ํž˜๋“  ๊ฒฝ์šฐ์™€ ๊ฐ™์ด ๋ณ€๋™์„ฑ์ด ์žˆ๋Š” ๋ฆฌ์Šคํฌ๋ฅผ ํ™•๋ฅ ๋ชจํ˜•์—์„œ ์ผ๋ จ์˜ ๋‚œ์ˆ˜๋ฅผ ๋ฐœ์ƒ์‹œ์ผœ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์œผ๋กœ ๋ณต์ˆ˜์˜ ํ•ด๋ฅผ ์ฐพ์•„ ์˜์‚ฌ๊ฒฐ์ •์„ ๋„์ถœํ•  ์ˆ˜ ์žˆ๋Š” ๋„๊ตฌ์ด๋‹ค. ํŠนํžˆ, ๋ถˆํ™•์‹คํ•œ ๋ณ€์ˆ˜๋“ค์„ ๊ณ ๋ คํ•  ๋•Œ, ์ˆ˜ํ•™์  ๋ฐฉ๋ฒ•์„ ์ ์šฉํ•˜๊ธฐ ์–ด๋ ค์šธ ๋•Œ ์‚ฌ์šฉํ•˜๊ธฐ ์œ ์šฉํ•˜๋‹ค(Shin, 2007). ์„ ํ–‰์—ฐ๊ตฌ ๋ถ„์„์„ ํ†ตํ•ด์„œ ์šฐ๋ฆฌ๋Š” ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด์„œ ๊ฐ€์žฅ ์ค‘์š”ํ•œ ๊ธฐํ›„์ธ์ž์ธ ์ผ์‚ฌ๋Ÿ‰๊ณผ ํŒจ๋„ ํšจ์œจ์„ ์ฃผ์š” ๋ณ€์ˆ˜๋กœ ํ•˜์—ฌ MCS๋ฅผ ์ ์šฉํ•˜์—ฌ ๋ฏธ๋ž˜์˜ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๋Š” ์—ฐ๊ตฌ๋ฅผ ํ•˜๊ณ ์ž ํ•œ๋‹ค.

3. ์—ฐ๊ตฌ๋Œ€์ƒํ˜„์žฅ ๋ฐ ์—ฐ๊ตฌ๋ฐฉ๋ฒ•๋ก 

ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ์˜ ์•ˆ์ •์ ์ธ ์ „๋ ฅ ๊ณต๊ธ‰๊ณผ ๊ฒฝ์ œ์ ์ธ ํƒ€๋‹น์„ฑ ํ™•๋ณด๋ฅผ ์œ„ํ•ด์„œ๋Š” ๋ฐœ์ „๋Ÿ‰์ด ๋งค์šฐ ์ค‘์š”ํ•˜๋‹ค. ์ƒˆ๋กœ์šด ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ๋ฅผ ๊ณ„ํšํ•˜๊ฑฐ๋‚˜ ์„ค์น˜ํ•˜๊ธฐ ์œ„ํ•ด์„œ๋Š” ๋งค์šฐ ์ •๋ฐ€ํ•œ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก๊ณผ์ •์ด ํ•„์š”ํ•˜๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก ๋ฐฉ๋ฒ•์€ ๊ณผ๊ฑฐ ๊ธฐ์ƒ์ž๋ฃŒ๋ฅผ ํ™œ์šฉํ•˜์˜€๋Š”๋ฐ, ํŠนํžˆ ์ผ์‚ฌ๋Ÿ‰, ์ผ์กฐ์‹œ๊ฐ„, ๊ธฐ์˜จ, ์šด๋Ÿ‰, ๊ฐ•์ˆ˜๋Ÿ‰ ๋“ฑ์˜ ์ •๋ณด๋ฅผ ์ฃผ๋กœ ๋ถ„์„ํ•˜์˜€๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” ์—ฐ๊ตฌ๋Œ€์ƒ ์ง€์—ญ์˜ ๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ๋ถ„์„ํ•˜์—ฌ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, MCS๋ฅผ ํ™œ์šฉํ•˜์—ฌ ์ผ์‚ฌ๋Ÿ‰๊ณผ ํŒจ๋„ํšจ์œจ์˜ ํ™•๋ฅ ๋ถ„ํฌ๋ฅผ ๋ถ„์„ํ•˜๊ณ , ๋‚œ์ˆ˜๋ฅผ ์ƒ์„ฑํ•˜์—ฌ ํ™•๋ฅ ๋ก ์ ์ธ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ณ ์ž ํ•œ๋‹ค.

3.1 ์ƒˆ๋งŒ๊ธˆ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์‚ฐ์ •

๋ณธ ์—ฐ๊ตฌ๋Š” ์ „๋ถ ๊ตฐ์‚ฐ์‹œ ์ƒˆ๋งŒ๊ธˆ์ง€์—ญ์— ์œ„์น˜ํ•œ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ๋ฅผ ์—ฐ๊ตฌ๋Œ€์ƒ ์ง€์—ญ์œผ๋กœ ํ•˜์˜€๋‹ค. Fig. 2๋Š” ์—ฐ๊ตฌ๋Œ€์ƒ ์ง€์—ญ์˜ ์œ„์น˜๋„์ด๋‹ค. ํƒœ์–‘๊ด‘ ์„ค์น˜ ๋ฉด์ ์€ 221,400 m2์ด๋‹ค. ํƒœ์–‘๊ด‘ ํŒจ๋„ ๊ฐœ์ˆ˜๋Š” 51,912์žฅ์œผ๋กœ ํƒœ์–‘๊ด‘ ํŒจ๋„ ์„ค์น˜ ๋ฉด์ ์€ 103,824 m2์ด๋‹ค. 2017๋…„์— ์‹œ์ž‘๋œ ํ”„๋กœ์ ํŠธ์˜ ์ด ํˆฌ์ž๋น„์šฉ์€ 431์–ต์›์ด๊ณ , ์šด์˜๊ธฐ๊ฐ„์€ ์ค€๊ณต๋œ 2018๋…„ ์ดํ›„ 20๋…„๋™์•ˆ์ด๋‹ค(KOEN, 2021). ๊ตฐ์‚ฐ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ๋Š” PE๋ถ€๋ ฅ์ฒด๋ฅผ ์‚ฌ์šฉํ•œ ๊ตฌ์กฐ์ด๋ฉฐ, ์ธ๋ฒ„ํ„ฐ๋Š” 1 MW๊ธ‰ 19๊ธฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ๊ตญ์‚ฐ ํŒจ๋„์„ ์‚ฌ์šฉํ•˜์˜€๋‹ค. Table 3๋Š” ํ”„๋กœ์ ํŠธ์˜ ๊ตฌ์ฒด์ ์ธ ์‚ฌ์–‘์„ ๋‚˜ํƒ€๋‚ธ ๊ฒƒ์ด๋‹ค.

ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ๊ณ„์‚ฐ์€ ํƒœ์–‘๊ด‘ ๋ฐœ์ „ํšจ์œจ, ๋Œ€์ƒ ์ง€์—ญ์˜ ๊ฒฝ์‚ฌ๋ฉด ์ผ์‚ฌ๋Ÿ‰, ํƒœ์–‘๊ด‘ ํŒจ๋„ ์„ค์น˜ ๋ฉด์ ์˜ ํ•จ์ˆ˜๋กœ ๊ตฌ์„ฑ๋˜์—ˆ์œผ๋ฉฐ, ๋ฐœ์ „๋Ÿ‰ ๊ณต์‹์€ Eq. (1)๊ณผ ๊ฐ™๋‹ค(Kim and Nam, 2010).

(1)
$E=\eta AQ$

์—ฌ๊ธฐ์„œ, ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์€ $E$[kWh]์ด๊ณ , ๋ฐœ์ „ํšจ์œจ์€ $\eta$์ด๋ฉฐ, ํƒœ์–‘๊ด‘ ํŒจ๋„ ์„ค์น˜ ๋ฉด์ ์€ $A$[m2], ๊ฒฝ์‚ฌ๋ฉด ์ผ์‚ฌ๋Ÿ‰์€ $Q$[kWh/m2]์ด๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ๊ณ„์‚ฐํ•˜๊ธฐ ์œ„ํ•ด ๊ธฐ์ƒ์ฒญ์—์„œ ์ œ๊ณตํ•˜๋Š” ์ข…๊ด€๊ธฐ์ƒ๊ด€์ธก(Automated Synoptic Observing System)์˜ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ–ˆ๋‹ค. ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ํ™•๋ณดํ•  ์ˆ˜ ์žˆ๋Š” ์ธ๊ทผ ์ง€์—ญ์ธ ์ „์ฃผ์‹œ ๊ณผ๊ฑฐ 10๋…„(2012-2021)๊ฐ„์˜ ์›” ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค(KMA, 2022). ํƒœ์–‘๊ด‘ ํŒจ๋„์˜ ๊ฒฝ์‚ฌ ๊ฐ๋„๋Š” 30ยฐ๋กœ ๊ฐ€์ •ํ•˜์˜€๊ณ , ๊ฒฝ์‚ฌ๋ฉด ์ผ์‚ฌ๋Ÿ‰์€ ์›” ํ‰๊ท  ๋น„์œจ์ธ 1.21๋กœ ์‚ฐ์ •ํ•˜์˜€๋‹ค(Jo et al., 2001). ๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ํƒœ์–‘๊ด‘ ๋ฐœ์ „ํšจ์œจ์— ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ๊ณผ ๊ธฐ๊ณ„์  ํšจ์œจ์„ ์ ์šฉํ•˜์—ฌ ๊ณ„์‚ฐํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์€ ์ œ์กฐ์‚ฌ ๋ณ„๋กœ ์ฐจ์ด๊ฐ€ ์žˆ์œผ๋‚˜, ์ตœ๊ทผ ์ œํ’ˆ๋“ค์€ ํŒจ๋„ ํšจ์œจ์ด 20 % ์ด์ƒ ํ™•๋ณดํ•˜๊ณ  ์žˆ๊ธฐ ๋•Œ๋ฌธ์— ๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ๊ธฐ์ค€(Baseline)์„ 20 %๋กœ ๊ฐ€์ •ํ•˜์˜€๋‹ค. ๊ธฐ๊ณ„์  ํšจ์œจ์€ 0.7์œผ๋กœ ์‚ฐ์ •ํ•˜์˜€๋‹ค(So et al., 2013).

Fig. 2. Floating Solar Power Location Map and Photo
../../Resources/KSCE/Ksce.2023.43.2.0249/fig2.png
Table 3. Project Descriptions of Floating Solar Power(KOEN, 2021)

Factor

Value

Site

Bieungdo-dong, Gunsan-si, Jeollabuk-do, South Korea

Facility

18.7 MW

SPC

P*** Company

Investment

43.1 billion won

Construction period

2017.02. - 2017.11

Complection

2018.10.30

Operation period

20 years after completion

Solar panel array

51,912 EA

Solar power station area

221,400 m2

Installed solar panel area

103,824 m2

Inverter

1 MW X 19 EA

Buoyant body type

25,925 EA (PE floating bodies)

3.2 ๋ชฌํ…Œ์นด๋ฅผ๋กœ ์‹œ๋ฎฌ๋ ˆ์ด์…˜(Monte Carlo Simulation: MCS)

MCS๋Š” ๋ณ€์ˆ˜๋ฅผ ๊ฐ€์ •ํ•œ ํ™•๋ฅ ๋ถ„ํฌ์— ๋งž๊ฒŒ ๋ฌด์ž‘์œ„ ํ‘œ๋ณธ์ถ”์ถœ์— ์˜ํ•ด ๋‚œ์ˆ˜๋ฅผ ๋ฐœ์ƒ์‹œ์ผœ ํ•จ์ˆ˜์˜ ๊ฐ’์„ ๋ถ„์„ํ•˜๋Š” ๊ธฐ๋ฒ•์ด๋‹ค. ๋ณธ ์—ฐ๊ตฌ์—์„œ ํƒœ์–‘๊ด‘ ํŒจ๋„์˜ ํšจ์œจ๊ณผ ์ผ์‚ฌ๋Ÿ‰์„ ๋ณ€์ˆ˜๋กœ ์‚ฌ์šฉํ•˜์—ฌ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ์˜ ์›”๋ณ„ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์˜ ํ™•๋ฅ ๋ถ„ํฌ๋Š” ์ •๊ทœํ™•๋ฅ ๋ถ„ํฌ๋กœ ๊ฐ€์ •ํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์€ ํŒจ๋„ ์ œ์กฐ์‚ฌ์˜ ๋ฐ์ดํ„ฐ์‹œํŠธ์— ์˜๊ฑฐํ•˜์—ฌ ํŒจ๋„ ํšจ์œจ์˜ Baseline์„ 20 %๋กœ ์‚ฐ์ •ํ•˜์˜€๋‹ค(Qcells, 2022). ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์€ ๋งŽ์€ ๋ฐœ์ „์„ ์ด๋ฃจ์–ด ๊ณผ๊ฑฐ 1์„ธ๋Œ€ ํŒจ๋„์€ 10 %์˜€์œผ๋‚˜, ์ตœ๊ทผ 2์„ธ๋Œ€ ํŒจ๋„์€ 20 %๋ฅผ ๋„˜์–ด ํ˜„์žฌ 25 %๊นŒ์ง€ ๊ธฐ์ˆ ์ด ๋ฐœ์ „ํ•˜๊ณ  ์žˆ๋‹ค(Jenkins and Ekanayake, 2017). ๊ฐ€๊นŒ์šด ๋ฏธ๋ž˜์— ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์ด ํš๊ธฐ์ ์œผ๋กœ ๊ฐœ์„ ๋  ๊ฒƒ์ด๋‹ค. ํŒจ๋„ ํšจ์œจ์— ์˜ํ–ฅ์„ ์ฃผ๋Š” ์š”์†Œ๋“ค์€ ์—ฌ๋Ÿฌ ๊ฐ€์ง€ ์žˆ๋‹ค. ํŠนํžˆ ๊ธฐํ›„์— ์˜ํ–ฅ์„ ๊ฐ€์žฅ ๋งŽ์ด ๋ฐ›๋Š”๋‹ค. ๊ธฐ์˜จ, ๋จผ์ง€, ์ฃผ๋ณ€ ์ง€์žฅ๋ฌผ, ๊ตฌ๋ฆ„, ๋ˆˆ, ๋น„, ๊ทธ๋ฆฌ๊ณ  ํƒœ์–‘์˜ ์œ„์น˜๋ณ€ํ™”์— ์˜ํ•ด์„œ ํšจ์œจ์ด ๋ณ€ํ™”ํ•œ๋‹ค. ๋˜ํ•œ ํŒจ๋„์˜ ์—ดํ™”๋กœ ์ธํ•˜์—ฌ ๋งค๋…„ 0.5 %์”ฉ ํšจ์œจ์˜ ์ €ํ•˜๊ฐ€ ๋ฐœ์ƒํ•œ๋‹ค(Yoo et al., 2009). ์ˆ˜์ƒํƒœ์–‘๊ด‘์˜ ๊ฒฝ์šฐ, ์ˆ˜์˜จ์˜ ์˜ํ–ฅ์œผ๋กœ ์—ฌ๋ฆ„์ฒ ์—๋Š” ๋ฐœ์ „ํšจ์œจ์ด ์œก์ง€๋ณด๋‹ค ์ƒ์Šนํ•˜๋Š” ์ด์ ์ด ์žˆ๋‹ค(Choi, 2014). ์ด๋Ÿฌํ•œ ์—ฌ๋Ÿฌ ๊ฐ€์ง€ ์š”์†Œ๋“ค๋กœ ์ธํ•˜์—ฌ ํŒจ๋„ ํšจ์œจ์ด ์ตœ์•…์˜ ๊ฒฝ์šฐ 10 %, ์ตœ์ƒ์˜ ๊ฒฝ์šฐ 30 %๊นŒ์ง€์˜ ๋ฒ”์œ„๋กœ ์ตœ๋Œ“๊ฐ’๊ณผ ์ตœ์†Ÿ๊ฐ’์„ ์„ค์ •ํ•˜๊ณ , ํ™•๋ฅ ๋ถ„ํฌ๋Š” ์ •๊ทœ๋ถ„ํฌ๋ฅผ ๋”ฐ๋ฅธ๋‹ค๊ณ  ๊ฐ€์ •ํ•˜์˜€๋‹ค. ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ์˜ ํ™•๋ฅ ๋ถ„ํฌํ•จ์ˆ˜๋Š” ์—ฌ๋Ÿฌ ์—ฐ๊ตฌ๋…ผ๋ฌธ์„ ์ฐธ๊ณ ํ•˜์—ฌ, ์ •๊ทœ๋ถ„ํฌ๋กœ ํ•˜์˜€๋‹ค(Al-Sumaiti et al., 2019; Dolatabadi, and Mohammadi-ivatloo, 2018; Kinanti et al., 2021). ์ผ์‚ฌ๋Ÿ‰์˜ ๊ฒฝ์šฐ, ๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ์›” ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ ํ‰๊ท  119.9791 kWh/m2๋ฅผ ๊ธฐ์ค€์œผ๋กœ ํ•˜์˜€๋‹ค. 2015๋…„ 11์›”์˜ ์›” ์ผ์‚ฌ๋Ÿ‰ 48.9973 kWh/m2์„ ์ตœ์†Ÿ๊ฐ’์œผ๋กœ ๊ฒฐ์ •ํ•˜์˜€์œผ๋ฉฐ, 2019๋…„ 5์›”์˜ ์›” ์ผ์‚ฌ๋Ÿ‰ 212.6335 kWh/m2์„ ์ตœ๋Œ“๊ฐ’์œผ๋กœ ๊ฒฐ์ •ํ•˜์˜€๋‹ค. ์›” ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ์˜ ์ •๊ทœ์„ฑ ๋ถ„ํฌ ๊ฒ€์ฆ์„ ์œ„ํ•ด Q-Q Plot ๋ฐฉ๋ฒ•์„ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์ž…๋ ฅ๋ฐ์ดํ„ฐ๋Š” 2012๋…„๋ถ€ํ„ฐ 2021๋…„๊นŒ์ง€์˜ ์›” ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ 120๊ฐœ์ด๋ฉฐ, Q-Q Plot์— ์˜ํ•˜์—ฌ ๊ฒ€์ฆํ•œ ๊ฒฐ๊ณผ ์ •๊ทœ๋ถ„ํฌ๋กœ ๊ฐ€์ •ํ•˜์—ฌ๋„ ๋ฌด๋ฆฌ๊ฐ€ ์—†๋‹ค๊ณ  ํŒ๋‹จํ•˜์˜€๋‹ค(Fig. 3). MCS๋Š” Oracle์‚ฌ์˜ Crystal Ball ํ”„๋กœ๊ทธ๋žจ์„ ์‚ฌ์šฉํ•˜์—ฌ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ถ„์„์„ ์‹ค์‹œํ•˜์˜€๋‹ค. Table 4๋Š” ๋ณ€์ˆ˜๋“ค์˜ ํ™•๋ฅ ๋ถ„ํฌ์™€ ์ตœ๋Œ€, ์ตœ์†Ÿ๊ฐ’์„ ์š”์•ฝํ•œ ๊ฒƒ์ด๋‹ค.

Fig. 3. Q-Q Plot Normality Test
../../Resources/KSCE/Ksce.2023.43.2.0249/fig3.png
Table 4. Probability Distribution of a Variable

Factor

Value

Minimum

Maximum

Distribution

Reference

Solar panel efficiency (%)

20

10

30

Normal

Qcells

(2022)

Insolation (kWh/m2)

119.9791

48.9973

212.6335

Normal

(Kinanti et al., 2021)

KMA

(2022)

4. ์—ฐ๊ตฌ๊ฒฐ๊ณผ ๋ถ„์„

๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ์ž๋ฃŒ๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ๋ฐœ์ „๋Ÿ‰์„ ๊ณ„์‚ฐํ•œ ๊ฒฐ๊ณผ, ํ†ต์ƒ 5์›”์˜ ๋ฐœ์ „๋Ÿ‰์ด ๊ฐ€์žฅ ๋†’์•˜๊ณ , 12์›”์ด ๊ฐ€์žฅ ๋‚ฎ์€ ๊ฒƒ์„ ๋ณด์—ฌ์ค€๋‹ค(Fig. 4). ๋˜ํ•œ ์›” ์ผ์‚ฌ๋Ÿ‰ ์ž๋ฃŒ๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์—ฐ๋ณ„ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ๊ณ„์‚ฐํ•ด๋ณด๋ฉด, ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์ด ๊ฐ€์žฅ ๋†’์€ ํ•ด๋Š” 2018๋…„์ด๊ณ , ๊ฐ€์žฅ ๋‚ฎ์€ ํ•ด๋Š” 2013๋…„์ด์—ˆ๋‹ค(Fig. 5). 2015๋…„ ์ดํ›„์˜ ์‹ค์ œ ๊ธฐ์ƒํ…Œ์ดํ„ฐ์ธ ์ผ์‚ฌ๋Ÿ‰์ž๋ฃŒ๋ฅผ ์ ์šฉํ•˜๋ฉด ๊ฑด์„ค ์ „ ์ˆ˜์ƒํƒœ์–‘๊ด‘ ์ถ”์ •์น˜(25.1 GWh)๋ณด๋‹ค ๋งŽ์ด ๋ฐœ์ „๋œ๋‹ค๊ณ  ์ถ”์ •ํ•  ์ˆ˜ ์žˆ๋‹ค.

Fig. 6๋Š” 10,000๋ฒˆ์˜ ๋ฐœ์ „๋Ÿ‰ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ถ„์„ ๊ฒฐ๊ณผ๋ฅผ ๋ณด์—ฌ์ค€๋‹ค. ๋ฐœ์ „๋Ÿ‰์˜ ์›” ํ‰๊ท ์€ 2.1 GWh์ด๋ฉฐ, ์›” ๋ฐœ์ „๋Ÿ‰์˜ ์ตœ์†Ÿ๊ฐ’๊ณผ ์ตœ๋Œ“๊ฐ’์€ ๊ฐ๊ฐ 0.3 GWh, 5.0 GWh์œผ๋กœ ์˜ˆ์ธก๋˜์—ˆ๋‹ค. MCS ๋ถ„์„์˜ ์ž์„ธํ•œ ๊ฒฐ๊ณผ๊ฐ’์€ Table 5์™€ ๊ฐ™๋‹ค.

๋ฐœ์ „ํšจ์œจ์€ ๊ธฐ๊ณ„์ ํšจ์œจ๊ณผ ํŒจ๋„ ํšจ์œจ์„ ๊ณ ๋ คํ•ด์„œ ๋ฐœ์ „๋Ÿ‰์„ ๊ณ„์‚ฐํ•˜์˜€๋‹ค. ๊ธฐ๊ณ„์  ํšจ์œจ์€ ํƒœ์–‘๊ด‘์‹œ์Šคํ…œ์˜ ์ „๊ธฐ์ , ๊ธฐ๊ณ„์  ์†์‹ค์— ์˜ํ•œ ํšจ์œจ์ด๋ฉฐ, ํŒจ๋„ ํšจ์œจ์€ ํŒจ๋„์ด ํƒœ์–‘๊ด‘์„ ์ „๊ธฐ๋กœ ์ „ํ™˜ํ•  ๋•Œ ์ ์šฉ๋˜๋Š” ๊ฒƒ์ด๋‹ค. ๊ธฐ๊ณ„์  ํšจ์œจ์€ 0.7๋กœ์„œ ํŒจ๋„์„ ์ œ์™ธํ•œ ํƒœ์–‘๊ด‘๋ฐœ์ „์‹œ์Šคํ…œ์˜ ์„ค๊ณ„๊ณ„์ˆ˜๋ผ๊ณ  ํ•œ๋‹ค(So et al., 2013). ๋ณธ ๋…ผ๋ฌธ์—์„œ๋Š” ํƒœ์–‘๊ด‘๋ฐœ์ „์— ์˜ํ–ฅ์„ ์ฃผ๋Š” ๋ฐœ์ „ํšจ์œจ ์ค‘ ๊ธฐ๊ณ„์  ํšจ์œจ์„ ์ œ์™ธํ•œ ํŒจ๋„ ํšจ์œจ์„ ๋ถ„์„ํ•˜์˜€๋‹ค.

๋ฐœ์ „๋Ÿ‰์„ ์‚ฐ์ •ํ•˜๋Š” ๊ณผ์ •์—์„œ ์ผ์‚ฌ๋Ÿ‰๊ณผ ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ ๋ณ€์ˆ˜๋“ค์ด ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋œ ๋ฐœ์ „๋Ÿ‰์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ์„ ์•Œ์•„๋ณด๊ธฐ ์œ„ํ•˜์—ฌ ๊ฐ ๋ณ€์ˆ˜๋“ค์˜ ๋ฏผ๊ฐ๋„๋ถ„์„์„ ์‹ค์‹œํ•˜์˜€๋‹ค. ๋ณธ ๋…ผ๋ฌธ์—์„œ MCS๋ฅผ ์ˆ˜ํ–‰ํ•˜๋Š”๋ฐ, 2๊ฐ€์ง€์˜ ์ž…๋ ฅ๋ณ€์ˆ˜(์ผ์‚ฌ๋Ÿ‰๊ณผ ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ)๋ฅผ ๊ฒฐ์ •ํ•˜๊ณ , ํ™•๋ฅ ๋ถ„ํฌํ•จ์ˆ˜๋ฅผ ์„ค์ •ํ•˜์˜€๋‹ค. ๊ทธ ๊ฒฐ๊ณผ๋กœ ๋ฐœ์ „๋Ÿ‰์˜ ๋ถ„ํฌ๊ฐ’์„ ์•Œ์•„๋ƒˆ์œผ๋ฉฐ, ๋ฐœ์ „๋Ÿ‰๋ถ„ํฌ๋ฅผ ์ •๋Ÿ‰ํ™”ํ•˜์˜€๋‹ค. ํ•˜์ง€๋งŒ, ๊ฐ ์ž…๋ ฅ๋ณ€์ˆ˜๊ฐ€ MCS๋กœ๋ถ€ํ„ฐ ์ถœ๋ ฅ๋œ ๊ฐ’(๋ฐœ์ „๋Ÿ‰)์˜ ๋ถˆํ™•์‹ค์„ฑ์— ์–ผ๋งˆ๋‚˜ ๊ธฐ์—ฌํ•˜๋Š”์ง€ ํ‰๊ฐ€ํ•˜๊ธฐ ์œ„ํ•ด์„œ ๋ฏผ๊ฐ๋„ ๋ถ„์„์„ ์‹ค์‹œํ•œ๋‹ค. ์ถœ๋ ฅ๊ฐ’(๋ฐœ์ „๋Ÿ‰)์˜ ๋ณ€ํ™”๋ฅผ ๊ฒฐ์ •ํ•˜๋Š” ๋ฐ ์žˆ์–ด์„œ ์ž…๋ ฅ๊ฐ’(์ผ์‚ฌ๋Ÿ‰๊ณผ ํŒจ๋„ ํšจ์œจ)์˜ ๋ณ€ํ™”๊ฐ€ ๋‘๊ฐ€์ง€ ์ž…๋ ฅ๊ฐ’ ์ค‘ ์–ด๋–ค ์š”์†Œ๊ฐ€ ๋” ์ค‘์š”ํ•˜๊ฒŒ ์ž‘์šฉํ•˜๋Š”์ง€ ์•Œ์•„๋‚ด๊ธฐ ์œ„ํ•ด์„œ MCS ์ˆ˜ํ–‰ ํ›„์— ๋ฏผ๊ฐ๋„๋ถ„์„์„ ๋ฐ˜๋“œ์‹œ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. ๋ฏผ๊ฐ๋„๋ถ„์„๊ฒฐ๊ณผ, ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์‚ฐ์ •์— ์˜ํ–ฅ์„ ์ฃผ๋Š” ๊ฒƒ์€ ์ผ์‚ฌ๋Ÿ‰์ด 58.8 %, ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์€ 41.2 % ์ˆœ์œผ๋กœ ๋ถ„์„๋˜์—ˆ๋‹ค(Fig. 7). ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์€ ์ผ์‚ฌ๋Ÿ‰์˜ ์˜ํ–ฅ์ด ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ๋ณด๋‹ค ๋” ํฌ๋‹ค๋Š” ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค.

๊ฑด์„ค ์ „ ์‚ฌ์—…์„ ๊ฐœ๋ฐœํ•˜๋Š” ํšŒ์‚ฌ์—์„œ ์˜ˆ์ƒํ•œ ์—ฐ๊ฐ„ ๋ฐœ์ „๋Ÿ‰์€ ๋ณธ ์—ฐ๊ตฌ์— ์˜ํ•œ MCS๋ถ„์„ ํ‰๊ท ๊ฐ’์ธ 25.3 GWh(์›”๊ฐ„ ํ‰๊ท ๋ฐœ์ „๋Ÿ‰ 2.1 GWh)๋กœ์„œ ๊ฑฐ์˜ ๋™์ผํ•˜๊ฒŒ ๋ฐœ์ƒํ•˜์˜€๋‹ค. Table 5๋Š” MCS๋ถ„์„์— ์˜ํ•œ ๊ฒฐ๊ณผ๊ฐ’๋“ค์ด๋‹ค. ํ•˜์ง€๋งŒ, ๋ถ„์„๊ฒฐ๊ณผ์— ์˜ํ•˜๋ฉด ๋ณธ ์—ฐ๊ตฌ๋Œ€์ƒ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ๋Š” ๋ฐœ์ „๋Ÿ‰์˜ ์›”๊ฐ„ ์ฐจ์ด๊ฐ€ ๋งŽ์ด ๋ฐœ์ƒํ•œ๋‹ค. ํŠนํžˆ, ์›”๊ฐ„๋ฐœ์ „๋Ÿ‰์ด ์ตœ๋Œ€์ธ ๋‹ฌ์€ 5์›”์ด๊ณ , ์ตœ์†Œ๋Š” 12์›”์ด๋ฉฐ, ๊ทธ ์ฐจ์ด๋Š” ์•ฝ 56 %์ด๋‹ค. ์—ฐ๊ฐ„ ๋ฐœ์ „๋Ÿ‰์˜ ์ฐจ์ด๋„ 2013๋…„(์ตœ์†Œ)์™€ 2018๋…„(์ตœ๋Œ€)๊ฐ„์— ์•ฝ 23 %์ด๋‹ค. ์œ„์˜ ๊ฒฐ๊ณผ๋ฅผ ํ†ตํ•ด์„œ ํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ์˜ ๋ฐœ์ „๋Ÿ‰์€ ์›”๋ณ„, ์—ฐ๋ณ„ ์ฐจ์ด๊ฐ€ ์‹ฌํ•˜๋‹ค๋Š” ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ์—ˆ๋‹ค. ๊ตฐ์‚ฐ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ์˜ ์›” ๋ฐœ์ „๋Ÿ‰์€ ํ‰๊ท  2.1 GWh์ด๋ฉฐ, ์ผ์‚ฌ๋Ÿ‰์˜ ๊ณ„์ ˆ์  ์ฐจ์ด๋กœ ์ธํ•ด ์‹œ๋ฎฌ๋ ˆ์ด์…˜์˜ ๊ฒฐ๊ณผ๋กœ์„œ 0.3~5.0 GWh์˜ ๋ฒ”์œ„์—์„œ ๋ฐœ์ „๋Ÿ‰์ด ํฌ๊ฒŒ ์ฐจ์ด๊ฐ€ ๋‚œ๋‹ค.

Fig. 4. Comparison of Estimated Solar Power Generation Per Month (2012-2021)
../../Resources/KSCE/Ksce.2023.43.2.0249/fig4.png
Fig. 5. Comparison of Estimated Solar Power Generation Per Year (2012-2021)
../../Resources/KSCE/Ksce.2023.43.2.0249/fig5.png
Fig. 6. MCS Results of Solar Power Generation
../../Resources/KSCE/Ksce.2023.43.2.0249/fig6.png
Fig. 7. Tornado Chart of Sensitivity Analysis for the Case Study
../../Resources/KSCE/Ksce.2023.43.2.0249/fig7.png
Table 5. MCS Results for the Case Study

Statistics

Predicted value

Statistics

Predicted value

Trials

10,000

Skewness

0.3238

Base case

2,116,619

Kurtosi

3.19

Mean

2,110,115

Coeff. of variation

0.2593

Median

2,083,182

Minimum

321,627

Standard deviation

547,238

Maximum

5,032,674

Variance

299,469,158,445

Mean Std. error

5,472

5. ๊ฒฐ ๋ก 

๋ณธ ์—ฐ๊ตฌ๋Š” ๊ณผ๊ฑฐ 10๋…„๋™์•ˆ์˜ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์—ฌ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ๊ณ„์‚ฐํ•˜๊ณ , ๋ถˆํ™•์‹ค์„ฑ์„ ์ค„์ด๊ธฐ ์œ„ํ•ด์„œ ํ™•๋ฅ ๋ก ์  ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ฐฉ๋ฒ•๋ก ์ธ MCS๋ฅผ ์‚ฌ์šฉํ•ด ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•˜๊ธฐ ์œ„ํ•ด์„œ, ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ์˜ˆ์ธกํ•œ ๊ธฐ์กด ์—ฐ๊ตฌ์™€ MCS๋ฅผ ์‚ฌ์šฉํ•œ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ๊ณ„์‚ฐ์— ๊ด€๋ จ๋œ ์—ฐ๊ตฌ๋ฅผ ์กฐ์‚ฌ, ๋ถ„์„ํ•˜์˜€๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” 18.7 MW ๊ทœ๋ชจ์˜ ์ƒˆ๋งŒ๊ธˆ์— ์œ„์น˜ํ•œ ๊ตฐ์‚ฐ ์ˆ˜์ƒํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ๋ฅผ ์—ฐ๊ตฌ๋Œ€์ƒ์œผ๋กœ ํ•˜์˜€๋‹ค. ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ๊ณ„์‚ฐ์— ์‚ฌ์šฉ๋œ ๊ณผ๊ฑฐ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋Š” ๊ธฐ์ƒ์ฒญ์ด ์ œ๊ณตํ•˜๋Š” ์›” ์ผ์‚ฌ๋Ÿ‰ ์ž๋ฃŒ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์›”๋ณ„ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์„ ์œ„ํ•ด ํŒจ๋„ ํšจ์œจ๊ณผ ์ผ์‚ฌ๋Ÿ‰์„ ์ •๊ทœ๋ถ„ํฌ์˜ ํ™•๋ฅ ๋ถ„ํฌ๋กœ ๊ฐ€์ •ํ•˜์—ฌ MCS๋ถ„์„์„ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค.

๋ณธ ์—ฐ๊ตฌ์˜ ๋ชฉ์ ์€ ๋งŽ์€ ์—ฐ๊ตฌ์—์„œ ๊ฒ€์ฆ๋œ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ฐฉ๋ฒ•์ธ MCS์„ ํ™œ์šฉํ•˜์—ฌ ๊ธฐํ›„๋ณ€ํ™”๋กœ ์ธํ•œ ๋ณ€๋™์„ฑ์˜ ๊ฒฐ๊ณผ๋ฅผ ์ •ํ™•ํ•˜๊ฒŒ ์˜ˆ์ธกํ•˜๊ธฐ ์–ด๋ ต๊ธฐ ๋•Œ๋ฌธ์— ํ™•๋ฅ ๋ชจํ˜•์—์„œ ์ผ๋ จ์˜ ๋‚œ์ˆ˜๋ฅผ ๋ฐœ์ƒ์‹œํ‚ค๋Š” ์‹œ๋ฎฌ๋ ˆ์ด์…˜๋ฐฉ๋ฒ•์œผ๋กœ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ๋ถ„์„์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ํ•ด์†Œํ•˜๊ณ ์ž ํ•˜์˜€๋‹ค. ๋ณธ ์—ฐ๊ตฌ ๊ฒฐ๊ณผ๋ฅผ ์ฐธ๊ณ ํ•˜์—ฌ ํƒœ์–‘๊ด‘ ๋ฐœ์ „์†Œ๋ฅผ ์œ„์น˜๋ฅผ ์„ ์ •ํ•˜๊ณ , ํƒ€๋‹น์„ฑ ์กฐ์‚ฌ๋ฅผ ํ•จ์— ์žˆ์–ด ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์˜ ๋ถˆํ™•์‹ค์„ฑ์„ ํ•ด์†Œํ•˜๋Š”๋ฐ ๋„์›€์ด ๋  ์ˆ˜ ์žˆ์„ ๊ฒƒ์ด๋‹ค.

๋ณธ ์—ฐ๊ตฌ์˜ ๊ฒฐ๊ณผ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ์ฒซ์งธ, ๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ๊ธฐ์ƒ์ฒญ ์ธก์ •์ž๋ฃŒ๋ฅผ ๋ถ„์„ํ•œ ๊ฒฐ๊ณผ, ๊ฐ€์žฅ ๋ฐœ์ „๋Ÿ‰์ด ๋งŽ์€ ๋‹ฌ์€ 5์›”(2.64 GWh), 3์›”(2.49 GWh) ์ˆœ์ด๋ฉฐ, 12์›”(1.69 GWh)๊ณผ 11์›”(1.73 GWh)์ด ์ ๋‹ค. ์ด๊ฒƒ์€ ์ผ์‚ฌ๋Ÿ‰์ด ๊ณ„์ ˆ๋ณ„๋กœ ์ฐจ์ด๊ฐ€ ์žˆ๊ธฐ ๋•Œ๋ฌธ์ด๊ณ  ๊ทธ ์ตœ๋Œ€์ฐจ์ด๋Š” ์•ฝ 64 %(0.94 GWh)์ด๋‹ค ์ฆ‰, ๋ฐœ์ „๋Ÿ‰์€ ๋ด„์— ์ตœ๋Œ€์ด๋ฉฐ, ๊ฒจ์šธ์— ์ตœ์†Œ๊ฐ€ ๋œ๋‹ค. ๋‘˜์งธ, MCS๋ฅผ ์‚ฌ์šฉํ•œ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๊ฒฐ๊ณผ๋กœ ํ‰๊ท  ์›” ๋ฐœ์ „๋Ÿ‰์€ 2.1 GWh์ด๋ฉฐ, ์ตœ์†Œ 0.3 GWh, ์ตœ๋Œ€ 5.0 GWh ์˜ˆ์ธก๋˜์—ˆ๋‹ค. ์ด๊ฒƒ์€ ์ผ์‚ฌ๋Ÿ‰๊ณผ ํŒจ๋„ ํšจ์œจ์˜ ๋ณ€๋™์œผ๋กœ ์ธํ•œ ๊ฒฐ๊ณผ์ด๋‹ค. ์ฆ‰, ์ผ์‚ฌ๋Ÿ‰๊ณผ ํŒจ๋„ ํšจ์œจ์ด ๊ฐ€์žฅ ํฐ ๊ฐ’์ด๋ฉด ๋ฐœ์ „๋Ÿ‰๋„ ์ตœ๋Œ€ ๋ฐœ์ƒํ•œ๋‹ค. ์…‹์งธ, ๋ฏผ๊ฐ๋„ ๋ถ„์„๊ฒฐ๊ณผ์— ์˜ํ•˜๋ฉด ํƒœ์–‘๊ด‘๋ฐœ์ „๋Ÿ‰์€ ํŒจ๋„ ํšจ์œจ ๋ณด๋‹ค ์ผ์‚ฌ๋Ÿ‰์˜ ์˜ํ–ฅ์„ ๋” ๋งŽ์ด ๋ฐ›๋Š” ๊ฒƒ์œผ๋กœ ๋ถ„์„๋˜์—ˆ๋‹ค. ์ด๊ฒƒ์€ ๋ฐœ์ „๋Ÿ‰์„ ์ตœ๋Œ€๋กœ ํ•˜๋ ค๋ฉด ํƒœ์–‘๊ด‘๋ฐœ์ „์†Œ ์œ„์น˜ ์„ ์ •์‹œ ์ผ์‚ฌ๋Ÿ‰์„ ์šฐ์„ ์ ์œผ๋กœ ๊ณ ๋ คํ•ด์•ผ ํ•œ๋‹ค๋Š” ๊ฒƒ์„ ์˜๋ฏธํ•œ๋‹ค. ํƒœ์–‘๊ด‘ ํŒจ๋„ ํšจ์œจ์€ ๊ธฐ์ˆ ์ด ๋ฐœ์ „ํ•˜๊ณ  ์žˆ์œผ๋ฏ€๋กœ ๊ฐ€๊นŒ์šด ๋ฏธ๋ž˜์—๋Š” ํŒจ๋„ ํšจ์œจ์ด ํš๊ธฐ์ ์œผ๋กœ ๊ฐœ์„ ๋  ๊ฒƒ์œผ๋กœ ๊ธฐ๋Œ€๋œ๋‹ค. ๋„ท์งธ, ๋ณธ ์—ฐ๊ตฌ์—์„œ ๊ณผ๊ฑฐ 10๋…„๊ฐ„์˜ ์‹ค์ œ ์ผ์‚ฌ๋Ÿ‰ ์ž๋ฃŒ๋ฅผ ์ ์šฉํ•˜์—ฌ ํƒœ์–‘๊ด‘ ๋ฐœ์ „๋Ÿ‰์„ ๋ถ„์„ํ•˜๋ฉด 2015๋…„ ์ดํ›„ ๋ฐœ์ „๋Ÿ‰์€ ๊ฑด์„ค ์ „ ์˜ˆ์ƒ๋ฐœ์ „๋Ÿ‰(์—ฐ๊ฐ„ 25.1 GWh)๋ณด๋‹ค ๋” ๋งŽ์Œ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค. ๊ธฐํ›„๋ณ€ํ™”๋กœ ์ธํ•ด ๋ฐœ์ „๋Ÿ‰ ๋ณ€๋™์„ฑ์ด ์ปค์ง€๊ณ  ์žˆ์ง€๋งŒ ์ถ”์„ธ์ ์œผ๋กœ ๋ฐœ์ „๋Ÿ‰์ด ์ฆ๊ฐ€ํ•  ๊ฐ€๋Šฅ์„ฑ์ด ์žˆ๋‹ค๊ณ  ํ•  ์ˆ˜ ์žˆ๋‹ค.

ํ•˜์ง€๋งŒ ๋ณธ ์—ฐ๊ตฌ๋Š” ๋‹ค์Œ๊ณผ ๊ฐ™์€ ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์— ์‚ฌ์šฉ๋œ ๊ธฐ์ƒ๋ฐ์ดํ„ฐ๋Š” ๊ณผ๊ฑฐ 10๋…„๊ฐ„ ์‹ค์ œ ๊ด€์ธก๋œ ๊ธฐ์ƒ์ฒญ ์ผ์‚ฌ๋Ÿ‰ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•˜์˜€์ง€๋งŒ ๋‹ค๋ฅธ ๊ธฐํ›„ ์š”์†Œ์ธ ๊ธฐ์˜จ, ์Šต๋„, ์šด๋Ÿ‰, ๊ธฐ์ƒ์ƒํƒœ ๋“ฑ์€ ๊ณ ๋ ค๋˜์ง€ ์•Š์•˜๋‹ค. ๋งŒ์•ฝ ์ข€ ๋” ๋งŽ์€ ๊ธฐํ›„์š”์†Œ๋“ค์ด ๊ณ ๋ ค๋˜์—ˆ์œผ๋ฉด ๋” ์ •๋ฐ€ํ•œ ๋ถ„์„์ด ๊ฐ€๋Šฅํ•  ๊ฒƒ์ด๋‹ค. ๋˜ํ•œ ์ตœ๊ทผ ์ด์Šˆ์ธ ๋ฏธ๋ž˜ ๊ธฐํ›„๋ณ€ํ™”์— ๋Œ€ํ•œ ๊ณ ๋ คํ•œ๋‹ค๋ฉด ๋ฏธ๋ž˜์˜ ๋ฐœ์ „๋Ÿ‰ ์˜ˆ์ธก์˜ ์‹ ๋ขฐ๋„๋ฅผ ๋†’์ผ ์ˆ˜ ์žˆ์—ˆ์„ ๊ฒƒ์ด๋‹ค.

๊ฐ์‚ฌ์˜ ๊ธ€

๋ณธ ์—ฐ๊ตฌ๋Š” ํ•œ๊ตญ์—ฐ๊ตฌ์žฌ๋‹จ์˜ ์ง€์›์„ ๋ฐ›์•„ ์ˆ˜ํ–‰๋œ ๊ณผ์ œ์ž…๋‹ˆ๋‹ค(NRF-2019R1C1C1010332).

๋ณธ ๋…ผ๋ฌธ์€ 2022 CONVENTION ๋…ผ๋ฌธ์„ ์ˆ˜์ •ยท๋ณด์™„ํ•˜์—ฌ ์ž‘์„ฑ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.

References

1 
Ahn, C. M., Joo, J. C., Kim, J. H., Choi, S. H., Jang, J. S. and Go, H. W. (2021). โ€œReview of installation status and major environmental issues of floating photovoltaic power plants (FPVs).โ€ Journal of Korean Society of Environmental Engineers, Vol. 43, No. 4, pp. 286-298 (in Korean).DOI
2 
Al-Sumaiti, A. S., Ahmed, M. H., Rivera, S., El Moursi, M. S., Salama, M. M. A. and Alsumaiti, T. (2019). โ€œStochastic PV model for power system planning applications.โ€ IET Renewable Power Generation, Vol. 13, No. 16, pp. 3168-3179. DOI: https://doi.org/10.1049/iet-rpg.2019.0345DOI
3 
Choi, H. C. (2014). An empirical study for operating characteristics analysis of 100 (kW) floated photovoltaic system, Masterโ€™s Dissertation, Hanyang University (in Korean).URL
4 
Dolatabadi, A. and Mohammadi-ivatloo, B. (2018). โ€œstochastic risk-constrained optimal sizing for hybrid power system of merchant marine vessels.โ€ IEEE Transactions on Industrial Informatics, Vol. 14, No. 12, pp. 5509-5517.DOI
5 
Gastli, A. and Charabi, Y. (2010). โ€œSolar electricity prospects in Oman using GIS-based solar radiation maps.โ€ Renewable and Sustainable Energy Reviews, Vol. 14, No. 2, pp. 790-797.DOI
6 
Jenkins, N. and Ekanayake, J. (2017). Renewable energy engineering, Cambridge University Press, Cambridge, UK. DOI: 10.1017/978 1139236256DOI
7 
Jeon, W. Y., Cho, S. M. and Cho, I. H. (2019). โ€œEstimating the uncertainty of net load of 2030 renewable generation.โ€ New & Renewable Energy, Vol. 15, No. 4, pp. 28-38 (in Korean).URL
8 
Jeong, J. H. and Chae, Y. T. (2018). โ€œImprovement for forecasting of photovoltaic power output using real time weather data based on machine learning.โ€ Journal of The Korean Society of Living Environmental System, Vol. 25, No. 1, pp. 119-125 (in Korean).URL
9 
Jo, D. K., Chun, I. S., Jeon, M. S., Kang, Y. H. and Auh, C. M. (2001). โ€œA study on the analysis of solar radiation on inclined surfaces.โ€ Journal of the Korean Solar Energy Society, Vol. 21, No. 3, pp. 19-24 (in Korean).URL
10 
Joo, I. H. (2014). โ€œFloating solar power generation technology overview and current status.โ€ Journal of Electrical World Monthly Magazine, pp. 37-41 (in Korean).URL
11 
Kang, G. W., Park, J. W. and Yang, H. S. (2022). โ€œA study on the prediction of solar photovoltaic power generation using multiple- linear regression.โ€ Proceedings of Symposium of the Korean Institute of Communications and Information Sciences, pp. 556-557 (in Korean).URL
12 
Kim, D. S. and Nam, S. W. (2010). โ€œA study on economic analysis of rural green-village planning using solar energy.โ€ Journal of the Korean Society of Agricultural Engineers, Vol. 52, No. 4, pp. 27-34 (in Korean).DOI
13 
Kim, D. S., Koo, S. M. and Nam, S. W. (2008). โ€œEstimating optimal probability distributions of daily potential photovoltaic power generation for development of rural green-village by solar energy - with area of seosan weather station -.โ€ Journal of The Korean Society of Agricultural Engineers, Vol. 50, No. 6, pp. 37-47 (in Korean).DOI
14 
Kim, J. G., Kim, D. H., Yoo, W. S., Lee, J. Y. and Kim, Y. B. (2017). โ€œDaily prediction of solar power generation based on weather forecast information in Korea.โ€ IET Renewable Power Generation, Vol. 11, No. 10, pp. 1268-1273. DOI: https://doi.org/10.1049/iet-rpg.2016.0698.DOI
15 
Kim, J. W. (2019). โ€œA solar power prediction scheme based on machine learning algorithm from weather forecasts.โ€ The Journal of Korean Institute of Information Technology, Vol. 17, No. 9, pp. 83-89 (in Korean).URL
16 
Kim, S. J., Yu, J. H., Ryu, D. G. and Chang, B. H. (2021). โ€œLong- term prediction of photovoltaic power generation based on machine learning for participation in the small-scale power brokerage market.โ€ Proceedings of the KIEE Conference, The Korean Institute of Electrical Engineers, pp. 573-574 (in Korean).URL
17 
Kim, S. W. and Ryu, J. H. (2020). โ€œA study on computing stochastic capacity of energy storage systems using Monte Carlo simulations.โ€ Korean Chemical Engineering Research, Vol. 58, No. 3, pp. 424-429 (in Korean).DOI
18 
Kinanti, S. P., Moeis, A. O. and Kaharudin, D. (2021). โ€œFeasibility analysis of a large scale floating photovoltaic power plant investment using financial modeling with the consideration of uncertainties factors.โ€ Proceedings of the Second Asia Pacific International Conference on Industrial Engineering and Operations Management, Surakarta, Indonesia, September 14-16, pp. 2238-2250.URL
19 
Ko, W. and Kim, J. H. (2020). โ€œClustering-based scenario reduction approaches for generation of adequate photovoltaic generation scenario.โ€ Proceedings of the KIEE Conference, The Korean Institute of Electrical Engineers, pp. 315-316 (in Korean).URL
20 
Korea Energy Agency (KEA) (2021). New & renewable energy statistics 2020 (2021 Edition), KEA, Ulsan, Korea (in Korean).URL
21 
Korea Meteorological Agency (KMA) (2022). Data, Available at: https://data.kma.go.kr/data/grnd/select AsosRltmList.do?pgmNo=36 (Accessed: July 1, 2022).URL
22 
Korea South-East Power Co. (KOEN) (2021). Public data, Available at: https://www.koenergy.kr/kosep /gv/nf/dt/nfdt20/main.do?gubun= KS5106 &menuCd=GV050 20306 (Accessed: July 1, 2022)URL
23 
Kwon, O. G., Choi, S. H., Jo, H. S. and Cha, H. J. (2022a). โ€œThe prediction of a floating photovoltaic generation utilizing RNN.โ€ The Transactions of The Korean Institute of Electrical Engineers, Vol. 71, No. 8, pp. 1126-1134 (in Korean).URL
24 
Kwon, T. H., Kim, J. Y., Kim, E. K. and Hong, S. K. (2022b). โ€œEffect on power generation of floating photovoltaic power system power by water level change.โ€ Journal of the Korean Solar Energy Society, Vol. 42, No. 2, pp. 13-21 (in Korean).DOI
25 
Lee, C. S. and Ji, P. S. (2015). โ€œDevelopment of daily PV power forecasting models using ELM.โ€ The Transactions of the Korean Institute of Electrical Engineers P. The Korean Institute of Electrical Engineers, Vol. 64, No. 3, pp. 164-168 (in Korean).DOI
26 
Lee, G. H. and Song, S. K. (2022). โ€œPredicting solar power generation based on deep neural network.โ€ Proceedings of the Korea Contents Association Conference, The Korea Contents Association, pp. 375-376 (in Korean).URL
27 
Lee, J. I., Park, W. K., Lee, I. W. and Kim, S. H. (2022). โ€œComparison of solar power prediction model based on statistical and artificial intelligence model and analysis of revenue for forecasting policy.โ€ Journal of IKEEE, Vol. 26, No. 3, pp. 23-31 (in Korean).URL
28 
Lee, K. H. and Kim, W. J. (2016). โ€œForecasting of 24_hours ahead photovoltaic power output using support vector regression.โ€ The Journal of Korean Institute of Information Technology, Vol. 14, No. 3, pp. 175-183 (in Korean).URL
29 
Lee, S. M. and Lee, W. J. (2016). โ€œDevelopment of a system for predicting photovoltaic power generation and detecting defects using machine learning.โ€ KIPS Transactions on Computer and Communication Systems, Vol. 5, No. 10, pp. 353-360 (in Korean).URL
30 
Lee, Y. R. and Kim, M. K. (2020). โ€œESS operation planning based on microgrid operating cost.โ€ Proceedings of the KIEE Conference, The Korean Institute of Electrical Engineers, pp. 387-388 (in Korean).URL
31 
Liew, N. and Lee, H. J. (2020). โ€œPerformance of a photovoltaic/ concentrated solar power hybrid system based on the splitting of the solar spectrum in Korea.โ€ Journal of the Korean Solar Energy Society, Vol. 40, No. 6, pp. 61-67 (in Korean).DOI
32 
Liu, W., Guo, D., Xu, Y., Cheng, R., Wang, Z. and Li, Y. (2018). โ€œReliability assessment of power systems with photovoltaic power stations based on intelligent state space reduction and pseudo-sequential Monte Carlo simulation.โ€ Energies, Vol. 11, No. 6, 1431.DOI
33 
Qcells (2022). Products. Available at: https://qcells.com/kr/get-started/ complete-energy-solution/ solar-panel (Accessed: July 1, 2022)URL
34 
Ryu, H. S., Lee, Y. R. and Kim, M. K. (2020). โ€œPredicting renewable energy generation using LSTM for risk assessment of local level power networks.โ€ The transactions of The Korean Institute of Electrical Engineers, Vol. 69, No. 6, pp. 783-791 (in Korean).DOI
35 
Saemangeum Development and Investment Agency (SDIA) (2022). Introduction of Saemangeum, Available at: http://www.saemangeum. go.kr (Accessed: July 1, 2022)URL
36 
Serttas, F., Hocaoglu, F. O. and Akarslan, E. (2018). โ€œShort term solar power generation forecasting: a novel approach.โ€ International Conference on Photovoltaic Science and Technologies (PVCon), Ankara, Turkey, pp. 1-4, DOI: 10.1109/PVCon.2018. 8523919.DOI
37 
Shin, D. B. (2007). โ€œForecasting exchange rates by Monte Carlo simulation.โ€ Journal of Industrial Economics and Business, Vol. 20, No. 5, pp. 2075-2093 (in Korean).URL
38 
Shin, D. H., Park, J. H. and Kim, C. B. (2017). โ€œPhotovoltaic generation forecasting using weather forecast and predictive sunshine and radiation.โ€ Journal of Advanced Navigation Technology, Vol. 21, No. 6, pp. 643-650. DOI: https://doi.org/10.12673/JANT.2017. 21.6.643 (in Korean).DOI
39 
So, J. H., Hwang, H. M., Jung, Y. S., Ko, S. W., Ju, Y. C. and Lim, H. M., (2013). โ€œDesign factor calculation and analysis of grid- connected photovoltaic system.โ€ Journal of the Korean Solar Energy Society, Vol. 33, No. 5, pp. 89-94 (in Korean).DOI
40 
Song, J. J., Jeong, Y. S. and Lee, S. H. (2014). โ€œAnalysis of prediction model for solar power generation.โ€ Journal of Digital Convergence, Vol. 12, No. 3, pp. 243-248 (in Korean).URL
41 
Song, K. J., Jeong, J. I., Moon, J. H., Kwon, S. C. and Kim, H. S. (2022). โ€œMuLti-Site Pv Power Forecasting Using TransGRU based on Delaunay Tirangulation.โ€ Proceedings of the KIEE Conference, The Korean Institute of Electrical Engineers, pp. 354-355 (in Korean).URL
42 
Yoo, S. P., Jin, J. S., Kim, H. K., Kim, Y. H., Jeong, S. D., Seo, Y. S. and Jeong, N. J. (2009). โ€œImproving the effectiveness of a photovoltaic system by water impinging jet on the surface of photovoltaic cells.โ€ Proceedings of the KSES 2009 Spring Annual Conference, pp. 241-244 (in Korean).URL