PERANCANGAN KONTROLER PID LEVEL DEAERATOR & KONDENSOR PADA STEAM POWER PLANT BERBASIS ALGORITMA GENETIKA

ABIDIN, ZAENAL (2021) PERANCANGAN KONTROLER PID LEVEL DEAERATOR & KONDENSOR PADA STEAM POWER PLANT BERBASIS ALGORITMA GENETIKA. S1 thesis, Universitas Mercu Buana Jakarta.

Preview	Text (HAL COVER) 01. Cover.pdf Download (287kB) | Preview
	Text (BAB I) 02. Bab 1.pdf Restricted to Registered users only Download (126kB)
	Text (BAB II) 03. Bab 2.pdf Restricted to Registered users only Download (668kB)
	Text (BAB III) 04. Bab 3.pdf Restricted to Registered users only Download (1MB)
	Text (BAB IV) 05. Bab 4.pdf Restricted to Registered users only Download (870kB)
	Text (BAB V) 06. Bab 5.pdf Restricted to Registered users only Download (120kB)
	Text (DAFTAR PUSTAKA) 07. Daftar Pustaka.pdf Restricted to Registered users only Download (191kB)
	Text (LAMPIRAN) 08. Lampiran.pdf Restricted to Registered users only Download (1MB)

Abstract

Make-up water system in the steam power plant works with closed circulation, but the circulation there is a loss of water mass which is partly caused by the blowdown system to maintain the steam quality and minor leakage that occur in the system. At the time of operation, the temperature of the fluid in the deaerator is around 150 °C with a pressure 8 kg/cm², meanwhile the temperature of the adding water is around 35-36 °C, when two fluids with large temperature deviations mixed, there can be potential for water hammer and thermal stress that can cause mechanical damage or catastrophic failure of the deaerator in the long term so that it can reduce the reliability and life time of the equipment. According to best practice, the make-up water system must be able to transfer to the hotwell with consideration of the temperature deviation between the condensate water and the addition water not being too large so that it can minimize water hammer. The process of fluid flow in the condensate water system is non-linear, so a reliable and robust controller is needed in regulating the water level in the deaerator and condenser. PID control has weaknesses related to tuning in determining the parameter values of Kp, Ki and Kd which is done manually, trial and error so that it takes time and requires experience in order to obtain optimal system performance. For this reason, it is necessary to optimize the PID controller tuning to obtain optimal Kp, Ki and Kd constants with genetic algorithms to fulfill the desired control specifications. Based on the Ziegler-Nichols oscillation method, for the deaerator level controller, the constant Kp = 53,8, Ki = 21,4 and Kd = 33,9, obtained an overshoot 18.1%, rise time 1150 seconds, steady-state error 0,3, meanwhile for the hotwell level controller the constant Kp = 97,3, Ki = 15,6 and Kd = 63,1, resulting maximum overshoot 25.7%, rise time 1120 seconds and steady-state error 0.07. The results of the genetic algorithm optimization for PID deaerator level control, obtained constant Kp = 97.3, Ki = 15.6 and Kd = 63.1 with fitness overshoot 2.1%, rise time 1190 seconds and steady-state error 0.01 for PID deaerator level control. Meanwhile, for hotwell level control PID, the limit constant Kp = 94.2, Ki = 20.7 and Kd = 71.5 with fitness overshoot 1.6%, rise time 1218 seconds and steady-state error 0.001. PID tuning with genetic algorithm produces faster overshoot and setting time than Ziegler-Nichols oscillation method, but has a slower rise time. For steady-state error both methods produce a value of <1% so that it meets the requirements for control system. Keywords : PID, Genetic algorithm, overshoot, rise time, error steady-state Sistem air penambah pada PLTU bekerja dengan sirkulasi secara tertutup, akan tetapi didalam proses sirkulasinya terjadi kehilangan massa air yang antara lain disebabkan oleh adanya sistem blowdown untuk menjaga kualitas uap dan adanya kebocoran minor yang terjadi didalam sistem. Pada saat unit pembangkit beroperasi, temperatur fluida didalam deaerator berkisar 150 °C dengan tekanan 8 kg/cm², sedangkan temperatur air penambah berkisar 35- 36 °C, ketika dua fluida dengan deviasi temperatur yang besar bertemu, dapat berpotensi terjadinya water hammer dan thermal stress yang dapat menyebabkan kerusakan mekanis atau catastrophic failure pada deaerator dalam jangka panjang sehingga dapat mengurangi keandalan dan life time peralatan. Sesuai best practice, sistem pengisian air dapat dialihkan ke hotwell dengan pertimbangan deviasi temperatur antara air kondensat dan air penambah tidak terlalu besar sehingga dapat menimilasir water hammer. Proses aliran fluida dalam condensate water system bersifat non-linier, sehingga diperlukan pengendali yang handal dan robust dalam mengatur ketinggian air di deaerator dan kondensor. Kendali PID memiliki kelemahan terkait tuning dalam penentuan nilai parameter Kp, Ki dan Kd yang dilakukan secara manual, trial and error sehingga memakan waktu dan membutuhkan pengalaman agar diperoleh performansi sistem yang optimal. Untuk itu, perlu dilakukan optimasi penalaan kontroler PID untuk mendapatkan konstanta Kp, Ki dan Kd yang optimal dengan metode algoritma genetika guna memenuhi spesifikasi kontrol yang diinginkan. Berdasarkan metode osilasi Ziegler-Nichols, untuk pengendali level deaerator didapatkan konstanta Kp = 53,8, Ki = 21,4 dan Kd = 33,9, menghasilkan maximum overshoot sebesar 18,1%, rise time sebesar 1150 detik, error steady-state sebesar 0,3, sedangkan untuk pengendali level hotwell didapatkan konstanta Kp = 28,2, Ki = 8,7 dan Kd = 22,8, menghasilkan maximum overshoot sebesar 25,7%, rise time sebesar 1120 detik dan error steady-state sebesar 0,07. Hasil optimasi algoritma genetika untuk PID pengendalian level deaerator, didapatkan konstanta Kp = 97,3, Ki = 15,6 dan Kd = 63,1 dengan fitness overshoot sebesar 2,1%, rise time sebesar 1190 detik dan error steadystate sebesar 0,01 untuk PID pengendalian level deaerator. Sedangkan untuk PID pengendalian level hotwell didapatkan konstanta Kp = 94,2 ; Ki = 20,7 dan Kd = 71,5 dengan maximum overshoot sebesar 1,6%, rise time sebesar 1218 detik dan error steady-state sebesar 0,001. Penalaan PID dengan algoritma genetika menghasilkan overshoot dan setling time yang lebih cepat dibandingkan dengan metode osilasi Ziegler-Nichols, akan tetapi memiliki rise time yang lebih lambat. Untuk error steady-state kedua metode tersebut menghasilkan nilai <1% sehingga memenuhi syarat untuk pengontrolan sistem yang baik. Kata Kunci : PID, Algoritma Genetika, overshoot, rise time, error steady-state

Actions (login required)

View Item