JP3847988B2 - Reactor power monitoring device - Google Patents

Description

ここに、φ_m （ｒ）は定常時における空間高次モード中性子束分布を、ｎ_m (t) はｍ次モード中性子束が時間的に変動する振幅関数をそれぞれ表わし、次式の固有値方程式を満たす。
【００７２】
【数２】

Ｌ（ｒ），Ｍ（ｒ）はそれぞれ時間に依存しない消滅（拡散、吸収等）演算子，生成（核分裂）演算子であり、λ_m は固有値を表わす。
【００７３】
また、固有値方程式で表わされる各固有値Ｋ_m 間には次式の大小関係が成立する。
【００７４】
【数３】

すなわち、（３）式より、最も低次モードの固有値λ₀ が最も大きく、定常時には、最も低次モードの固有値λ₀ のみが最終的に残るために、最も低次モードを基本モードと指称する。
【００７５】
基本モードの時間的変動は良く知られた動特性方程式で表わされる。ｍ次モード中性子束（以下、ｍ＞０のモードを高次モードと称する。）の時間的変化は、次の（４）および（５）式の動特性方程式で表わされる。
【００７６】
【数４】

【数５】

（４）式におけるΔρm はｍ次モードの基本モードに対する未臨界度であり、次式で定義される。
【００７７】
【数６】

【００７８】
一方、各高次モード間については、以下のような直交関係が成立している。
【００７９】
【数７】

【００８０】
しかし、原子炉のような大型炉心では動的な高次モード分布歪曲が無視でき、また一般に炉心管理は１／４炉心対称で行なわれていることから、高次モード間に直接直交関係がほぼ満足される。
【００８１】
【数８】

【００８２】
（７）式および（８）式において、φ_m+（ｒ）は随伴中性子束を意味するが、以下では一群の中性子束を考えるので、φ_m+（ｒ）は中性子束分布φ_m （ｒ）に一致している。すなわち、次式で表わされる。
【００８３】
【数９】

【００８４】
（１）式と（９）式から、ある特定のモード成分、例えばｎ次モードの時間的変化だけを抽出するためには、中性子束全体にｎ次モード中性子束分布を炉心全体に亘って掛けてやればよい。すなわち、
【数１０】

【００８５】
したがって、フィルタとして注目する中性子束モード分布を用いれば良い。ただ、測定されるのは中性子束分布Φ（ｒ，ｔ）ではなく、炉心内に多数（例えば典型的な１１０万ＫＷｅ級ＢＷＲで１７２箇所）存在するＬＰＲＭ信号を用いることになる。すなわち、炉心内のある場所におけるＬＰＲＭ信号をＬＰＲＭ（ｒ，ｔ）として、その周辺における高次モード分布を用いることにより、
【数１１】

【００８６】
また、（１１）式において、重みとして場所に寄らない一定値を用いるのがＡＰＲＭ信号である。このＡＰＲＭ信号には全ての中性子束モードの時間変化が含まれていることになる。このように、実際の炉心内における過渡的な中性子束変動には、多くの高次モードが一般的には含まれていると考えられるが、炉心一体振動では基本モードが、領域振動では方位角方向１次モードの変動成分が支配的であることが、解析的にも実機振動データの評価結果からも確認されている。従来はＡＰＲＭ信号により炉心安定性を評価していたが、より正確には基本モードで重み付けされた信号を用いる必要がある。
【００８７】
このような原理に基づくと、フィルタ係数として高次モード分布を用いるのが最適である（米国特許第５４０６５９８号明細書参照）が、現時点においては高次モード分布を逐一求めて、フィルタ係数として利用できるようにはなっていない。
【００８８】
上記理由から、本発明の本実施形態は、より簡易的な手法を工夫して用いることにより、高次モード分布を用いる手法と同等以上の効果をもたらし、さらには実装がより簡単な原子炉出力監視装置を提案する。
【００８９】
まず、実炉心状態における高次モード分布が求まらない場合には、裸の均質円筒型炉心の中性子束分布φ_lmn （ｒ，θ，ｚ）を用い、３方向のインデックスを（ｌ，ｍ，ｎ）とすると、
【数１２】

【００９０】
炉心一体振動に関係するモードである基本モードでは、ｍ＝０，ｌ＝ｎ＝１に対応する。また、領域振動に関係するモードである周方向１次モードでは、ｌ＝ｍ＝ｎ＝１に対応する。
【００９１】
したがって、ｚ（軸）方向モードについては基本モードと周方向１次モードは一致する。また、高次モードｍ≧１では方位角θに対して周期的となるため、ｍ個の直線状の零点が加わってくる。したがって、周方向１次モード（ｍ＝１）では互いに直交する２つの異なる分布が、同じ固有値に対して存在する（すなわち、固有値が２重に縮退している）。
【００９２】
今回注目しているのは、炉心断面方向における中性子束分布の安定性であるので、フィルタ係数としては、ｍ次ベッセル関数Ｊ_m （ｘ）を用いる。基本モードはｍ＝０，ｎ＝１に対応するので、フィルタ係数としてＪ₀ （Ｘ_0,1 ｒ／Ｒ）を用いれば炉心安定性に関係するモードの、周方向１次モードではｍ＝ｎ＝１であるので、フィルタ係数としてＪ₁ （Ｘ_1,1 ｒ／Ｒ)cosθを用いれば領域安定性に関するモードをそれぞれ選択的に抽出することが可能である。
【００９３】
あくまでもｍ次ベッセル関数は裸の均一炉心体系におけるモード分布であるため、実機炉心のモード分布には一致しないが、炉心管理は１／４炉心対称に行なわれている。このため、中性子束分布は炉心断面についてはｍ次ベッセル関数に比較的近い分布を示す（図６参照）ため、安定性監視の目的には充分な精度が得られると考えられる。
【００９４】
図６において右図は代表的なＢＷＲの実機炉心における実際の周方向一次モード分布を示す。図６の右図に対し、左図は、均質炉本体における周方向一次モード分布に相当する一次ベッセル関数の分布を示している。ベッセル関数分布は、実機炉心における周方向一次モード分布に近似していることがわかる。
【００９５】
ただし、周方向１次モードは重根であるため、Ｊ₁ （Ｘ_1,1 ｒ／Ｒ)cosθと、それに直交するＪ₁ （Ｘ_1,1 ｒ／Ｒ)sinθの２種類が存在する。したがって、仮に実際の振動モードが、用意されたフィルタ分布と直交する分布であれば、そのフィルタでは振動モードが抽出できないことになってしまう。実際に発生する領域振動モードは正確には予測はできないことから、フィルタとして互いに直交する２つの周方向一次モード、すなわちＪ₁ （Ｘ_1,1 ｒ／Ｒ)cosθとＪ₁ （Ｘ_1,1 ｒ／Ｒ)sinθの２種類を予め用意しておけばよい。
【００９６】
ここで、便宜的に基本モード分布を重みとして用いて求められた安定性監視用信号をＲＰＲＭ０、互いに直交する周方向１次モードの重みを用いて求められた安定性監視用信号を、それぞれＲＰＲＭ１／ＲＰＲＭ２と呼ぶことにする。このようなフィルタによる各種ＲＰＲＭ信号生成手順を図７に示す。
【００９７】
図７ではＡＰＲＭ信号（平均化信号）を用いる際の重み分布の概念図を示している。現行用いられているＡＰＲＭ信号は全て同じ重み係数を使用して求められているが、本発明における安定性監視用信号（ＲＰＲＭ信号）は監視対象とする高次モードに対応した重み分布を用いて求めることを示している。すなわち、基本モードに対応した安定性監視用信号ＲＰＲＭ０では重みとして基本モードを、周方向一次モードに対応した安定性監視用信号ＲＰＲＭ１／ＲＰＲＭ２ではそれぞれ互いに直交した周方向一次モード分布を重みとしてそれぞれ用いる。
【００９８】
特に、空間的に位相差の生じる領域振動時には、各ＬＰＲＭ信号間で位相差があるために、従来のＡＰＲＭ信号では互いにキャンセルしてしまい、信号感度が低下してしまう。それに対してＲＰＲＭ１／ＲＰＲＭ２信号では、フィルタに使用した重み係数により、位相差が補正されてキャンセリングが起こらず、高い感度のＲＰＲＭ信号が得られる。これらの事情を図８に示す。
【００９９】
図８には、従来の方法では、各ＬＰＲＭ信号を全て同じ重み係数を用いてＡＰＲＭ信号を求めているため、求められたＡＰＲＭ信号はここのＬＰＲＭ信号が互いにキャンセリングしてしまい、ＡＰＲＭ信号自体が減衰してしまう。これに対し、本発明におけるＲＰＲＭ信号では、個々のＬＰＲＭ信号のキャンセリングが生じないために、着目する高次モードの応答が抽出できることが示されている。
【０１００】
次に、フィルタを用いた振動監視の例を、３次元動特性解析コードを用いた模擬振動データに適用した例を説明する。
【０１０１】
まず、図９に炉心一体振動に適用した例を示す。基本モードの安定性監視用信号ＲＰＲＭ０により振動現象を的確に捕えているが、周方向１次モードの安定性監視用信号ＲＰＲＭ２（ＲＰＲＭ１も同じ）には殆ど振動成分が含まれていない。すなわち、安定性監視用信号ＲＰＲＭ０と安定性監視用信号ＲＰＲＭ１／ＲＰＲＭ２との応答の差異から、この振動現象が炉心一体振動であることが判別できることがわかる。
【０１０２】
同様に、領域振動へ適用した例を図１０に示す。ＲＰＲＭ０／ＲＰＲＭ１に較べてＲＰＲＭ２の振動が有意に大きく、また周期も明確になっている。このことから、ＲＰＲＭ２の重み係数の分布に一致した方向（この例では振動の中心線が１３５度）の領域振動であることを判別できる。ＲＰＲＭ信号の重み係数分布とＲＰＲＭ信号応答の関係について図１１に示す。
【０１０３】
図１１では図１０に使用した解析結果を使用しているので、ＲＰＲＭ２が領域振動の振動方向に一致した重み係数を使用しており、ＲＰＲＭ１はそれに直交した重み係数を使用している。ここで、両者の中間の重み係数、すなわち、ＲＰＲＭ１／ＲＰＲＭ２から共に４５度ずれた重み分布を用いて求めたＲＰＲＭ信号を仮にＲＰＲＭ３と名付けると、ＲＰＲＭ３の振幅は両者の中間的な値になっていることがわかる。
【０１０４】
すなわち、領域振動の方向（振動の中心線、すなわち領域振動として現れている空間高次モードの零点の向き）は重み係数の分布を介してＲＰＲＭ信号の振幅（感度）と密接に関係していることを示している。重み係数としてＪ₁ （Ｘ_1,1 ｒ／Ｒ)sinθを用い、ここで重み係数の符号（正負）の替わる中心線角度、すなわち方位角θを０度から１８０度まで５度間隔で変えたときの、ＲＰＲＭ信号の平均振幅の変化を図１２に示す。
【０１０５】
振動の中心線（１３５度）と方位角θが一致するときがＲＰＲＭ信号の振幅は最大、すなわち感度が最大となり、逆にそれに直交する角度（４５度）で感度は最小となっている。したがって、予め互いに直交する重み係数を２種類（Ｊ₁ （Ｘ_1,1 ｒ／Ｒ)cosθとＪ₁ （Ｘ_1,1 ｒ／Ｒ)sinθ）を用意しておけば、少なくとも片方のＲＰＲＭ信号の振幅は、図１２の網かけを施した領域にあることになる。このことから、領域振動を監視するためには互いに直交する周方向１次モード分布に類似した重み係数を２種類用意しておくことが必要にして充分な条件である。
【０１０６】
また、図１２からもわかるように重み係数の分布角度を可変にして（あるいはこの図に示してあるように角度を少しずつ変えた重み係数を多数個用意して、それぞれのＲＰＲＭ信号を求める）、ＲＰＲＭ信号の振幅が最大になる角度から、領域振動の方向が判別できる。しかし、互いに直交する２種類の重み係数を用いた、ＲＰＲＭ１信号とＲＰＲＭ２信号の位相図の挙動からも、領域振動の方向を推定することが可能である。その事情を図１３に示す。
【０１０７】
図１３においては、互いに直交する２種類の重み係数を用いたＲＰＲＭ１信号とＲＰＲＭ２信号の応答の、重み係数の中心線の方向に対する変化を、両信号の位相図で示したものである。この位相図の応答が最大となる方位角θから、振動の中心線の方向がわかる。すなわち、ＲＰＲＭ１がＲＰＲＭ２に比べて十分大きい場合（図の右下方向）には、ＲＰＲＭ１に対応した高次モード不安定性が支配的であるため、振動の方向はその高次モード分布に一致する。逆に図１３の左上方向なら、振動の方向は高次モード分布に直交した方向となる。これらの中間的な方向（対角線近傍）であれば、両分布の中心線の中間に振動の方向があることになる。
【０１０８】
すなわち、ＲＰＲＭ１／ＲＰＲＭ２のいずれかのＲＰＲＭ信号の振幅が有意に偏っていれば、領域振動の方向はその振幅の大きなＲＰＲＭ信号の重み係数分布の中心線の方向により近く、同じ程度であれば両者の中間的な方向であることが推定できる。したがって、ＲＰＲＭ１信号とＲＰＲＭ２信号の相図を常に監視して表示装置２２で画面表示することも、領域安定性監視にとっては有効な手段である。
【０１０９】
ここで、両ＲＰＲＭ信号の振幅が同じ程度である場合に、領域振動であるか炉心一体振動であるかの判定は、図９にも示したように、ＲＰＲＭ０信号の振幅と比較すればよい。すなわち、ＲＰＲＭ０信号の振幅がＲＰＲＭ１／ＲＰＲＭ２信号の振幅に較べて有意に大きければ炉心一体振動であり、逆に小さければ領域振動であると判定できる。
【０１１０】
したがって、各ＲＰＲＭ信号から求められる安定度の指標である減幅比だけでなく、各信号の振幅、周期、および位相空間上での相互関係が出力振動現象の監視にあたっては必要であり、本発明はそれらの総合的判断に基づく振動現象の判定法を提供する手段に関するものである（図４および図５）。
【０１１１】
さらに、高次モードまでも含めた原子炉出力の安定性の監視には以上述べた手法が応用できる。すなわち、重み係数として以下の値を用いる。
【０１１２】
【数１３】

ここで用いている記号は（１２）式と同じである。
【０１１３】
この任意の高次モードまで拡張された重み係数を用い、ある時間毎に方位角θを回転させて、最も出力振幅の大きくなる方位角を設定する。この方位角は使用する高次モード、すなわち（１３）式における（ｍ，ｎ）の値毎に異なってくる。設定された最適方位角を（１３）式に代入した重み係数を用いて、各高次モードに対するＲＰＲＭ信号を求める。ここでは、方位角が最適化されているのでＲＰＲＭ信号は１種類で済む。
【０１１４】
求められたＲＰＲＭ信号が各高次モードの時間的な情報を持っているので、この信号から減幅比、振幅、周期等の安定度監視パラメータを抽出する。図１４にその監視例を示す。
【０１１５】
基本モードから任意の高次モードについて、上述した手法に基づいて、各モードの方向（図右上）、各モードの時間応答を表わすＲＰＲＭ信号の減幅比と周期（周波数）（図右下）、および各ＲＰＲＭ信号の応答波形と振幅の推移を、各モードについて表示する。
【０１１６】
この例では、基本モード成分、すなわち炉心一体振動成分が主要成分であり、例えば１次モード成分は僅かしか含まれていないが、それ以外にある高次モード成分も含まれており、減幅比自体はその高次モード成分の方が高いことがわかる。
【０１１７】
もう１つの監視項目として、各ＲＰＲＭ信号のスペクトルが挙げられる。図１５には、各ＲＰＲＭ信号の応答波形と振幅の推移とスペクトル（自己パワースペクトル密度：ＡＰＳＤ）を表示した例を示す。
【０１１８】
ＡＰＳＤのピーク高さおよびその鋭さから、各モードの安定度を判定することができるが、さらにスペクトルの基本周波数に対して、その整数倍の位置にピーク、すなわち高調波ピークが現われているか否かにより、そのモード成分が振動として支配的に発達したものであるかどうかが判定できる。
【０１１９】
例えば、図１６には発達した出力振動時のＲＰＲＭ信号のＡＰＳＤ（自己パワースペクトル密度）を示すが、基本周波数（約０．３５Ｈｚ）の２倍、３倍の周波数の高調波が明確に現われているが、図１５のスペクトルにはそれらは現われていない（ここで１Ｈｚ以上で現われているピークは安定性に関係しない他の理由によるものである）。
【０１２０】
したがって、ここで示した例においては比較的高い減幅比を示しており、高次モード成分も含まれてはいるが、振幅はピーク間値でも高々３％であり、安定度が悪化して発振に至る時点で顕著となる非線形高調波成分も含まれていないことから、炉心全体では比較的安定であり、特に安全に関わるような状態ではないことがわかる。振幅が一番大きな基本モード成分と、安定度自体は悪化しているＮ次モード成分について重点的に監視していけばよい。
【０１２１】
ここでは、炉心内で利用できるＬＰＲＭ信号を全て用いた監視を行なっているので、表示法としてはこれらの変化成分を立体的に、実時間で表示することにより、空間的な変動を視覚的に把握することが可能である。
【０１２２】
【外１】

【０１２３】
【数１４】

と規格化された偏差値を求め、実時間でそれらの変動を炉心内における検出器の配置に対応して３次元的な監視装置画面に表示する。図１７はその表示例を示す。
【０１２４】
図１７の表示例では振動周期の１周期分、約２秒間のプロットを示してあるが、実際にはこのような立体図がほぼ実時間でアニメーションとして表示装置２２で表示される。また、図１７の下には各ＬＰＲＭ信号振幅の等高線図を表示してあるが、振幅の値に応じて表示する色を変えてやれば、炉心内における出力変動の空間的な挙動を表示装置１９で視覚的にも把握し易くなる。
【０１２５】
また、ＬＰＲＭ信号の代りに、（１３）式で最適化された重みで各ＬＰＲＭ信号を修正した、いわゆる高次モードに対するＬＰＲＭ信号を用いれば、重みで用いた高次モード成分の変動をアニメーションとして表示することが同様に可能である。
【０１２６】
以上は、炉心内に満遍なく配置されたＬＰＲＭを全て、あるいは大多数を使用して監視する手法であるが、より少ない限られたＬＰＲＭ信号を用いて領域安定性をより簡易的に監視する手法について以下に説明する。
【０１２７】
例えば図６に示したように周方向１次モードは、炉心周辺部が高い分布となっているため、領域安定性に関する高次モードの炉心中心部のＬＰＲＭ信号への寄与は周辺部に較べて小さいと考えられる。したがって、従来通りＡＰＲＭ信号によって炉心安定性を監視するのであれば、領域安定性監視には周辺部のＬＰＲＭ信号だけを利用すればよいことになる。さらに、炉心中心からほぼ等距離にあるＬＰＲＭ検出器のＬＰＲＭ信号を使用すれば、重み係数としては同じ値を近似的に用いることができる。
【０１２８】
より少ない限られたＬＰＲＭ信号を用いた、簡易炉心安定性監視用に使用するＬＰＲＭ信号の炉心内配置の例を図１８に示す。この例では炉心安定性監視用に炉心中心部の５本のＬＰＲＭ信号を、領域性安定性監視用に炉心周辺部の１０本のＬＰＲＭ信号、計１５本のＬＰＲＭ信号を用いている。炉心安定性監視用の信号重み係数として均一に０．２、領域安定性監視用の重み信号としては±０．２を用いる。
【０１２９】
例えば、図１８において領域安定性監視用ＬＰＲＭ信号の中で、（３〜７）のＬＰＲＭに０．２を、（８〜２）のＬＰＲＭに−０．２の重みを使用した信号をＲＰＲＭ１、（１〜５）のＬＰＲＭに０．２を、（６〜１０）のＬＰＲＭに−０．２の重みを使用した信号をＲＰＲＭ２とする。また、各ＬＰＲＭ信号は、全てのＤＣ値により規格化されているとする。この規格化により、重み係数として均一な値を使用し、各信号のＤＣ値からの偏差成分を抽出するわけである。
【０１３０】
図１９は炉心安定性監視へ、図２０には領域安定性監視へ、それぞれ適用した例を示す。炉心一体振動時にはＲＰＲＭ０の振幅が卓越しており、領域振動時にはＲＰＲＭ１またはＲＰＲＭ２のいずれかが卓越しているので、前記ＬＰＲＭ信号を用いた場合と同様に安定性監視が行なわれる。但し、安全には他のモードからの影響が削除できないので、減幅比のみによる監視は困難であり、振幅と併せた評価が必要になってくる。
【０１３１】
図１９および図２０は、安定性予測装置２１におけるプロセス制御計算機１９に内蔵された３次元安定性解析コードを用いて、振動現象を模擬し、その時の核計装信号（ＬＰＲＭ）に本発明で提案した重みを用いて求めた、監視信号計算装置２５および安定性監視装置２６からのＲＰＲＭ０（基本モードに対応する炉心安定性監視信号）信号とＲＰＲＭ１（周方向１次モードに対応する領域安定性監視信号）信号の応答を示している。
【０１３２】
以上は、時系列データからの逐次的な安定性監視手法に関する内容であるが、さらに安定性予測装置２１による物理モデルに基づいた安定性予測機能について図２１を参照して説明する。
【０１３３】
まず、安定性予測機能における計算の流れについて、安定性監視装置２６における安定性監視の計算の流れと比較して図２１に示す。
【０１３４】
運転員（ユーザー）が、所定の運転条件を有する任意の運転点における安定性予測を行なう場合に、その予測したい運転条件（出力と冷却剤流量等）を安定性予測装置２１に入力する（図２１のステップＳ３０参照）。安定性予測装置２１は、プロセス制御計算機１９に記憶された上記入力時点における最新の炉心状態情報データ（炉心管理データ；図２１のＤ１参照）にアクセスして、出力分布を受け取る（ステップＳ３１参照）。そして、安定性予測装置２１は、受け取った出力分布から、まず最も出力の高い燃料集合体を選択して、チャンネル安定性を求める集合体とする。
【０１３５】
次に、安定性予測装置２１は、上記出力分布に応じて、熱水力的に条件の近い燃料集合体同士を幾つかまとめて、炉心安定性解析用の少数のチャンネルグループに分割する（ステップＳ３２参照）。
【０１３６】
同様に、安定性予測装置２１は、上記出力分布および安定性監視装置２６による安定性監視機能によりそれぞれ推定された振動中心線（Ｌ１参照）と、高次モード重み係数に基づいて、燃料集合体同士を幾つかまとめて領域安定性解析用の少数のチャンネルグループに分割する（ステップＳ３３参照）。
【０１３７】
また、領域安定性解析には高次モード未臨界度が必要であるが、現在の炉心監視システムでは高次モードの評価を行なっていないため、高次モード未臨界度は、予め出力分布の関数Ｇとして求めておき、監視システムから得られた出力分布を用いて計算される（ステップＳ３４参照）。
【０１３８】
高次モード未臨界度を求めるための手順は、次の通りである。すなわち、ある燃料集合体の出力分布をＰｒn として、次の出力分布の二乗平均Ｒを求める。
【０１３９】
【数１５】

【０１４０】
同じく、その燃料集合体の炉心中心からの距離に対する二乗平均ＲＬを求める。
【０１４１】
【数１６】

【０１４２】
すると、高次モード未臨界度は以上の２つの指標Ｒ、ＲＬの差の関数として表される。また、予測したい運転条件に対する未臨界度として、運転条件（出力Ｐ、冷却剤流量Ｆ）の関数Ｈを予めフィッティングしておく。すると、予測したい運転条件に対する未臨界度Δρは、
【数１７】

で与えられる。
【０１４３】
この他にも運転条件によって変わる変数、例えばバイパス部流量、チャンネル軸方向出力分布等を運転条件の関数として予め与えておき、炉心管理データ（炉心状態情報データ）の与えられている運転条件と、安定性を予測したい運転条件との状態の偏差より、予測運転状態における変数値を設定する。また、運転条件から定常のヒートバランスを用いて推定できる変数値、例えば炉心圧力、炉心入入口温度等は、ヒートバランス物理モデル（図２１のＨ１参照）より計算して設定される（ｆ）。
【０１４４】
以上のように、予測運転状態におけるチャンネル分割数、未臨界度の値および出力分布の値等を含む熱水力条件が求められたら、安定性予測装置２１は、そのチャンネル分割数、未臨界度、出力分布等の値を安定性解析用物理モデル（図２１におけるＭ１参照）に取り込み（ステップＳ３５参照）、炉心安定性、チャンネル安定性および領域安定性の解析をそれぞれ行ない（ステップＳ３６参照）、チャンネル・炉心安定性の減幅比および固有周波数をそれぞれ計算し（ステップＳ３７参照）かつ領域安定性の減幅比および固有周波数をそれぞれ計算する（ステップＳ３８参照）。
【０１４５】
また、図２１に示したように、安定性予測装置２１で求められた予測モデルは常時プラント（すなわち、原子炉１０）から得られる炉心状態情報データを含む安定性情報を利用することができるため（ステップＳ３９参照）、予測処理実行時においては、学習機能を実行することにより、図４のステップＳ２５ａ〜２５ｄおよびステップＳ２６ａ〜２６ｇに略対応するステップＳ４１〜Ｓ４７に示された安定性監視処理により、ＬＰＲＭデータを含むプラント時系列データより得られた安定性と、予測機能から得られた安定性とのバイアスを補正することができる（ステップＳ４０参照）。学習による調整パラメータとしては、炉心・領域安定性では出力分布、ボイド反応度係数が、炉心安定性単独では再循環系の伝達特性、領域安定性では高次モード未臨界度が考えられる。
【０１４６】
例えば、上記予測機能がユーザーにより明示的に起動された場合、および運転条件あるいは過渡事象等により、現状の運転点からある程度状態が変化した場合に、安定性監視装置２６による安定性監視機能と安定性予測装置２１による予測機能が自動的に起動されることにより、上記学習機能はある程度の頻度をもって行なわれる。ボイド反応度係数は炉心安定性と領域安定性の両者に効いてくるパラメータであるが、再循環系の伝達特性は炉心安定性にのみ、未臨界度は領域安定性のみに効いてくるパラメータであることから、まずボイド反応度係数により両安定性の調整学習を行ない、その後に再循環系の伝達特性により炉心安定性の調整学習を、未臨界度により領域安定性の調整学習を行なう。以上の学習機能は、図４の符号２１の安定性予測装置で実現される。
【０１４７】
すなわち、本発明においては中性子束分布の基本モードと高次モード分布を考慮した重み係数を用いた新たな監視信号を監視信号計算装置２５で求め、その信号応答から安定性監視装置２６で減幅比、振幅および位相関係を各係数について求め、それらから炉心内における空間的な振動現象を網羅的、総合的に判定することが可能である。また、安定性予測装置２１では最新の運転管理データを用い、物理モデルに基づいた安定性予測により、運転状態が変化した場合の炉心安定性の変化が予測でき、より柔軟な運用が可能となる。
【０１４８】
この実施例の原子炉出力監視装置では、従来のそのままアナログ信号を平均化するＡＰＲＭ信号に較べて、より正確に、また従来信号では検出困難な出力変動も含めて検出、判定が可能であり、安全性の向上と稼働率の向上を図ることができる。
【０１４９】
【発明の効果】
以上に述べた通り、本発明に係る原子炉出力監視装置においては、炉心内の原子炉炉心状態の中性子束の基本モード分布および互いに直交する高次モード分布に対応する重み係数を用いて従来通りの検出信号をフィルタ処理して安定性監視信号を生成することにより、従来のＡＰＲＭ信号では検出が困難となる原子炉出力変動を上記従来通りの検出信号を用いて正確に監視することができ、安全性の向上と稼働率の向上を図ることができる。
【０１５０】
本発明の原子炉出力監視装置は、従来の原子炉出力監視手法では見落しがちな原子炉出力変動を正確に監視できる一方、運転状態が変化した場合の炉心安定性の変化が予測でき、より柔軟な運用が可能となる。
【図面の簡単な説明】
【図１】本発明に係る原子炉出力監視装置の一実施形態を示す構成図。
【図２】図１に示された原子炉の炉心内における中性子束モニタの配置例を示す炉心の平断面図。
【図３】図２に示された中性子束モニタの炉心軸方向（鉛直方向）における配置例を示す平断面図。
【図４】本発明に係る原子炉出力監視装置における安定性監視および安定性予測の処理の流れを概略的に示すフローチャート。
【図５】本発明に係る原子炉出力監視装置の機能モジュールを示すブロック図。
【図６】本発明に係る原子炉出力監視装置において、フィルタ重みとして用いる出力分布例を示すもので、ベッセル関数分布と実機周方向一次モード分布の比較図。
【図７】本発明に係る原子炉出力監視装置において導入された新しいＲＰＲＭ信号の作成法を示す図。
【図８】本発明に係る原子炉出力監視装置によるフィルタの効果を示す図。
【図９】本発明に係る原子炉出力監視装置における出力安定性監視例を示すもので、炉心一体振動時の適用結果を示すＲＰＲＭ信号の応答例を示す図。
【図１０】本発明に係る原子炉出力監視装置における出力安定性監視例を示すもので、領域振動時の適用結果を示すＲＰＲＭ信号の応答例を示す図。
【図１１】本発明に係る原子炉出力監視装置における出力安定性監視例である重み係数の分布とＲＰＲＭ応答の関係を示す図で、領域振動時において、高次モードに対する重み係数を炉心方位角方向に回転させた際の、監視信号の応答の違いを比較した図。
【図１２】本発明に係る原子炉出力監視装置における出力安定性監視例である重み係数中心線角度とＲＰＲＭ振幅の関係を示す図で、領域振動時において、互いに直交した重み係数で求めた監視信号の位相線図上における応答と、領域振動の中心線方向との対応を示した図。
【図１３】本発明に係る原子炉出力監視装置における出力安定性監視例を示すもので、ＲＰＲＭ１／ＲＰＲＭ２相図と領域振動の方向を示す図。
【図１４】本発明に係る原子炉出力監視装置において、各モード成分を、減幅比、固有周波数および監視信号生波形とその振幅の観点から監視する高次モード監視例を示す図。
【図１５】本発明に係る原子炉出力監視装置において、各高次モード成分を、監視信号生波形とその振幅および監視信号のスペクトルの観点から監視する高次モードスペクトル監視例を示す図。
【図１６】本発明に係る原子炉出力監視装置において、発達した振動時にスペクトルに現われる高調波ピークの例を示す図。
【図１７】本発明に係る原子炉出力監視装置における規格化されたＬＰＲＭ信号変動の立体的表示例を示すもので、監視信号、この場合は平滑化されたＬＰＲＭ信号の変動分を、各ＬＰＲＭ検出器の炉心内配置に対応した位置に、逐次的にアニメーション表示した例を示す図。
【図１８】簡易炉心運転性監視用のＬＰＲＭ検出器の炉心内配置例を示す図。
【図１９】本発明に係る原子炉出力監視装置における簡易炉心安定性監視例であり、炉心一体振動に適用した結果を示す、簡易安定性監視信号による炉心安定性監視例の図。
【図２０】本発明に係る原子炉出力監視装置における簡易炉心安定性監視例てあり、領域振動に適用した結果を示す、簡易安定性監視信号による領域安定性監視例の図。
【図２１】本発明に係る原子炉出力監視装置において、安定性予測機能における計算の流れを、逐次的安定性監視機能との計算の流れと併せて示したフローチャート。
【符号の説明】
１０ 原子炉
１１ 原子炉圧力容器
１２ 炉心
１３ 中性子束検出装置（中性子束測定手段）
１４ 燃料集合体
１５ 中性子束検出器（ＬＰＲＭ）
１６ 制御棒
１７ 炉心現状データ測定器
１８ Ａ／Ｄ変換器
１９ プロセス制御計算機（中性子束分布算出手段）
２０ 入出力装置（入出力手段）
２１ 安定性予測装置（安定性予測手段）
２２ 表示装置
２４ 高次モード計算装置
２５ 監視信号計算装置（フィルタ計算装置）
２６ 安定性監視装置（安定性監視手段）[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a reactor power monitoring device that monitors the reactor power and its power distribution, and more particularly, to a reactor power monitoring device that is used to monitor the output fluctuation of a boiling water reactor.
[0002]
[Prior art]
In the core of a boiling water reactor (BWR) as a light water reactor, a large number of neutron flux detectors are arranged as neutron flux measuring means in the core as a nuclear reactor instrumentation system, The power distribution in the axial direction and radial direction of the core is monitored and instrumented.
[0003]
In the neutron flux detector, about one fuel assembly is arranged in every 16 BWR cores, and in each neutron flux detector, four neutron flux detectors are arranged in the vertical direction. These neutron flux detectors are called local output region monitors (LPRM). For example, 43 × 4 = 172 (channels) are arranged in the core of a 1100 MWe class BWR.
[0004]
The output signals (LPRM signals) from each neutron flux detector are averaged by an average output region monitor (APRM) every 20 or so. For example, in an 1100 MWe class BWR, 6 channels of average output area monitors are provided, and 6-channel APRM signals are output. These APRM signals and LPRM signals are all analog signals.
[0005]
In the conventional BWR, the APRM signal from the average power range monitor (APRM) is monitored, and when the APRM signal exceeds a certain set point, a trip signal such as a scram signal is output, and the reactor is operated in a dangerous state of the reactor. This makes it possible to operate the reactor stably.
[0006]
In particular, in the case of a boiling water reactor, in order to avoid an unstable phenomenon of the reactor power in the core, a reactor operation restriction region is provided, and the operation of the reactor within the operation restriction region is avoided. In the unlikely event that the operation of the nuclear reactor enters the operation restriction region, measures are taken such that a control rod selected in advance is inserted to lower the reactor output and to exit the operation restriction region.
[0007]
The reactor operation restriction area is obtained from the result of stability analysis using an analysis code in advance by a process control computer, but the stability can be evaluated from the fluctuation of the neutron flux detector signal (LPRM signal) in the core. A stability monitor has also been developed. However, the stability monitor evaluates the stability using the averaged APRM signal, or only evaluates the individual LPRM signal, and can accurately evaluate the stability of the power fluctuation of the reactor. Can not. The stability of reactor power fluctuations depends on the complex spatial dependence in the core, and has been observed in many overseas nuclear plants these days.
[0008]
[Problems to be solved by the invention]
The APRM signal can average the LPRM signals evenly, so that the reactor power distribution can be detected when the power fluctuates in the entire core, but the reactor core varies locally, In the case of fluctuation with a phase difference, the APRM signal is sucked after the fluctuation amount is canceled out by averaging the LPRM signal, which may make it difficult to detect the reactor power distribution.
[0009]
As an example of local fluctuations in the core, a so-called channel oscillation phenomenon in which a fuel assembly that is severe in thermohydraulicity generates an oscillation phenomenon called density wave oscillation can be considered. Even if this oscillation phenomenon is diffused by neutron flux oscillation, there is a possibility that it will fluctuate only within a relatively narrow range.
[0010]
In addition, as an example of the fluctuation having a spatial phase difference, there is an oscillation phenomenon called a region vibration that vibrates with a phase difference of 180 ° at each target position in the core. This oscillation phenomenon has actually been observed in some foreign nuclear plants. For example, in the region vibration observed at the CAORSO plant in Italy, the maximum amplitude of APRM was about 10% at the maximum, whereas in LPRM having the largest amplitude, an amplitude of 60% was observed. This is probably because half of the core vibrates with a phase difference of 180 ° from each other, so that the maximum value and the minimum value of LPRM are averaged simultaneously, and canceling occurs between them.
[0011]
In monitoring the core stability, the stability of the core state is estimated by obtaining the reduction ratio, the period of vibration, the amplitude, etc. representing the stability from the normal APRM signal. However, there is a possibility that the core stability cannot be accurately detected by monitoring only the APRM signal during the region vibration.
[0012]
In the reactor stability monitoring device, several LPRM signals at different locations in the core are selected and estimated in the same manner as the APRM signal. However, logic for processing and judging a plurality of LPRM signals Has not been determined, it is not used for detection of region vibration.
[0013]
As a method of solving the problem that the core stability cannot be accurately detected during the region vibration, a reference LPRM signal is selected in advance based on the signal dispersion value, and the phase difference of the selected LPRM signal is sequentially determined. By obtaining a neutron flux detection signal with higher sensitivity than simple averaging, or presuming a spatial higher-order mode distribution that may vibrate in advance and weighting this mode distribution The present applicant has proposed a technique used as (see US Pat. No. 5,406,598 and Japanese Patent Application No. 5-209717).
[0014]
The present invention has been made in consideration of the above-described circumstances, and can accurately monitor the fluctuations in the output of the reactor using conventional detection signals, thereby improving safety and operating rate. An object of the present invention is to provide a reactor power monitoring apparatus capable of
[0015]
Another object of the present invention is to improve the proposed reactor power monitoring method to an online power monitoring method for online monitoring, and to monitor reactor power fluctuations that are often overlooked by the conventional power monitoring method. It is in providing the reactor power monitoring apparatus which can do.
[0016]
[Means for Solving the Problems]
In order to solve the above-described problem, a reactor power monitoring apparatus according to the present invention includes a plurality of neutron flux measuring means arranged in the core of a nuclear reactor and the neutron fluxes as described in claim 1. Monitoring signal calculation device comprising a process control computer for estimating the power state in the core based on the neutron flux detection signal measured by the measuring means, and a filter for extracting the characteristics of the detection signal fluctuation based on the neutron flux detection signal And a stability monitoring device that evaluates the stability of the core from the value of the neutron flux detection signal subjected to the necessary filter in this monitoring signal calculation device, A display device for displaying the evaluation result, and a stability prediction device for predicting an arbitrary operation state using the function of the process control computer. The monitoring signal calculation device comprises the neutron flux detection signal as a fundamental mode distribution of neutron flux in a reactor core state and Higher order mode distribution orthogonal to each other A monitoring signal is generated by filtering with a corresponding weighting factor, and the stability monitoring device is configured to estimate and evaluate the stability of the core based on the monitoring signal obtained by the filtering process. Yes.
[0020]
Also, In order to solve the above-described problems, a reactor power monitoring apparatus according to the present invention is Claim 2 As described above, the stability monitoring device When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, Two types of output signals from the monitoring signal calculation device are evaluated on the phase plane, and the center line of the region stability is estimated from the difference in amplitude.
[0021]
further, In order to solve the above-described problems, a reactor power monitoring apparatus according to the present invention is Claim 3 As described above, the monitoring signal calculation device When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, With the obtained filter, an arbitrary higher-order mode distribution weight is rotated in the core azimuth direction, and the center line of the higher-order mode distribution is estimated by the stability monitoring device from the response amplitude. .
[0022]
Furthermore, In order to solve the above-described problems, a reactor power monitoring apparatus according to the present invention is Claim 4 As described in When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, The stability monitoring device determines the reduction ratio, vibration period, and amplitude from the output signal from the monitoring signal calculation device obtained by applying the optimum weighting coefficient in the core azimuth direction, and evaluates any higher-order mode stability. Is configured to Claim 5 As described in Said Monitoring signal calculator As a method for evaluating the stability of the core stability monitoring device using an index different from the reduction ratio based on the value of the neutron flux detection signal with the required filter applied While determining the spectral density, Said The stability monitoring device is configured to evaluate the importance of the contribution to the stability of higher-order modes by checking the presence or absence of harmonic components corresponding to integer multiples of the fundamental frequency in the obtained spectral density. .
[0023]
Moreover, in order to solve the above-described problems, the reactor power monitoring apparatus according to the present invention Claim 6 As described in Said Outputs the basic mode and higher-order mode output signals obtained by the monitor signal calculation device, and outputs the reduction ratio, vibration period, amplitude and spectrum obtained by evaluating this output signal with the stability monitor device. It is configured to display on the screen together with the response on the display device.
[0024]
On the other hand, the reactor power monitoring device according to the present invention is to solve the above-described problems, Claim 7 As described in Said The monitoring signal calculation unit obtains the smoothed fluctuation of the neutron flux detection signal from the neutron flux measuring means arranged in the core, and the fluctuation monitoring unit inputs this fluctuation and sequentially measures the core. The system is configured to evaluate the spatial fluctuation of the output in the core by displaying the animation on the three-dimensional coordinates corresponding to the position on the display device.
[0025]
On the other hand, the reactor power monitoring apparatus according to the present invention is to solve the above-described problems, Claim 8 As described in Said The monitoring signal calculation device multiplies the smoothed fluctuations of the neutron flux detection signal from the neutron flux measuring means arranged in the core by the weight distribution corresponding to each higher order mode, and monitors the higher order mode fluctuations as a stability monitoring device. Thus, the spatial variation of the higher-order mode in the core is evaluated by displaying the animation on the display device on the three-dimensional coordinates corresponding to the measurement position of the core sequentially.
[0026]
Furthermore, the reactor power monitoring apparatus according to the present invention is to solve the above-described problems, Claim 9 As described in Said The monitoring signal calculation device When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, When there are few neutron flux detection signals that can be used, a simple output signal is calculated using the neutron flux detection signal in the peripheral part that is equidistant from the core center, and the stability monitoring device The reduction ratio and the amplitude are obtained, and the region stability is evaluated.
[0027]
Furthermore, the reactor power monitoring apparatus according to the present invention is to solve the above-described problems, Claim 10 As described in Said The monitoring signal calculation device When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, Each neutron flux detection signal is normalized by its direct current value, the absolute value is equal, and a simple output signal is calculated using a weighting factor with half the sign and positive and negative. The monitoring device is configured to obtain the reduction ratio and amplitude and evaluate the region stability.
[0028]
Furthermore, in order to solve the above-described problem, a reactor power monitoring apparatus according to the present invention includes: Claim 11 As described in Said The monitoring signal calculation device When calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other, While outputting the output signal obtained using two kinds of weighting coefficients whose center lines for inverting the sign of the weighting coefficient are orthogonal to each other, the stability monitoring device obtains the reduction ratio and the amplitude from this output signal, thereby stabilizing the region. It is configured to simply evaluate the property.
[0029]
Moreover, in order to solve the above-described problems, the reactor power monitoring apparatus according to the present invention Claim 12 As described in The process control computer or the stability prediction device is: Stability analysis model based on physical model based on latest core management data With The stability prediction device is configured to predict the core / channel / region stability at an arbitrary operating point.
[0030]
Furthermore, the reactor power monitoring apparatus according to the present invention is to solve the above-described problems, Claim 13 As described in Said The stability prediction device is a parameter that changes as the operating point changes, and a function that is not physically modeled is approximated by a function within the operating point range that changes in advance, and as a deviation from the latest operation management data, It is configured to perform parameter evaluation at an operating point for predicting stability.
[0031]
Furthermore, the reactor power monitoring apparatus according to the present invention is to solve the above-described problems, Claim 14 As described in Said By comparing the core / regional stability reduction ratio predicted by the stability prediction device with the core / regional stability reduction ratio during operation evaluated by the monitoring signal calculation device and the stability monitoring device, The stability prediction apparatus is an apparatus in which the accuracy of the prediction function is improved by learning.
[0032]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of a reactor power monitoring apparatus according to the present invention will be described with reference to the accompanying drawings.
[0033]
FIG. 1 is a block diagram showing the overall configuration of a reactor power monitoring apparatus according to the present invention. Reference numeral 10 in the figure denotes a boiling water reactor (BWR) as a light water reactor. In this nuclear reactor 10, a reactor core 12 is stored in a reactor pressure vessel 11, and a plurality of neutron flux detectors 13 are arranged in the reactor core 12 as neutron flux measuring means.
[0034]
As shown in FIG. 2, the neutron flux detector 13 is arranged as a nuclear reactor instrumentation system in the core 12 at a ratio of about one for every 16 fuel assemblies 14, for example, during the power operation of the nuclear reactor. The nuclear power is monitored by monitoring the nuclear power output and the axial and radial power distribution of the core 12. For example, 43 neutron flux detectors 13 are arranged in the core 12 of an 1100 MWe class BWR and 43 in the core 12 of an 1300 MWe class BWR. As shown in FIG. 3, each neutron flux detector 13 has four neutron flux detectors 15 arranged in the vertical direction as local output region monitors (LPRMs). , B, C, D. In FIG. 2, reference numeral 16 denotes a control rod.
[0035]
The core 12 of the nuclear reactor 10 is usually provided with more than 100 neutron flux detectors 15. For example, 43 × 4 = 172 (channel) neutron flux detectors exist in the core of a 1100 MWe class BWR, and an output signal is taken out from each neutron flux detector 15 as an LPRM signal as an analog signal.
[0036]
The LPRM signals from these neutron flux detectors 15 are grouped by, for example, about every 20 signals by an average output region monitor (APRM), and are averaged evenly to become APRM signals.
[0037]
Further, as shown in FIG. 1, a reactor core current data measuring device 17 is installed inside the nuclear reactor 10, and the reactor core current data measuring device 17 provides a core state quantity other than reactor power, coolant flow rate, and cooling. A core operating state signal (plant state quantity measurement signal), which is core current data (process data) such as material temperature and pressure or control rod insertion amount, is measured. This core operating state signal is also an analog signal.
[0038]
Then, output data signals from the neutron flux detectors 15 of the neutron flux detector 13 and the core current data measuring device 17 are input to the A / D converter 18. The A / D converter 18 samples the input / output data signals (each LPRM signal, each core operation state signal, and each APRM signal) which are analog signals, and converts them into digital data.
[0039]
Each LPRM signal, which is a nuclear instrumentation signal digitized by the A / D converter 18, a core operating state signal, which is a plant state quantity measurement signal, and each APRM signal are respectively neutron flux data, core data, and APRM data as neutrons. It is output to a process control computer (process computer) 19 which is a bundle distribution calculation means and a monitor signal calculation device 25 which is a filter calculation device. In the monitoring signal calculation device 25, a filter is obtained based on the digital signal.
[0040]
On the other hand, the process control computer 19 is started periodically or at a timing based on an operator's request sent from the input / output device 20, and calculates the neutron flux distribution of the fundamental mode in the core 12 at the time of starting. ing. The process control computer 19 calculates the neutron flux distribution in the fundamental mode of the current core 12 and corresponds to a monitoring function of the reactor core state.
[0041]
In addition to the monitoring function, the process control computer 19 has a prediction function. In this prediction function, the core state designated by the operator is calculated and predicted based on the result of the latest monitoring function, and the basic in the predicted core state is calculated. The neutron flux distribution of the mode is calculated.
[0042]
On the other hand, the current calculation result of the neutron flux distribution of the core 12 calculated by the process control computer 19 and the predicted calculation result of the neutron flux distribution of the core 12 are output via the input / output device 20 which is an input / output means, to the operator. While being notified, it is input to the stability predicting device 21 and the core state is displayed on the display device 22. The stability predicting device 21 may be built in the process control computer 19.
[0043]
In addition, the process control computer 19 has a built-in high-order mode calculation device 24 as high-order mode calculation means. The higher-order mode calculation device 24 may be provided independently of the process control computer 19. The high-order mode of the neutron flux calculated by the process control computer 9 is calculated, and the calculation result of the indoor high-order mode of the neutron flux is input to the monitoring signal calculation device 25 as filter calculation means. Note that the high-order mode calculation method by the high-order mode calculation device 24 is specifically described in, for example, the document “Detailed Numerical Computation” written by Tohito Togawa, Kyoritsu Publishing Co., Ltd.
[0044]
On the other hand, each LPRM signal (digital LPRM signal), which is a nuclear instrumentation signal data sampled and digitized by the A / D converter 18, is input to the monitoring signal calculation device 25, and based on each input LPRM signal. While an optimum filter based on the phase difference and amplitude difference between the signals is obtained, the monitor signal calculation device 25 is also calculated by the higher-order mode calculation device 24 and a filter corresponding to the extracted spatial higher-order mode is also calculated. The That is, the monitoring signal calculation device 25 obtains a filter according to the core state or the variation characteristic of each actually measured LPRM signal.
[0045]
Further, in the monitoring signal calculation device 25, each input actual measured LPRM signal is subjected to filtering (averaging operation) by being applied to a filter based on a phase difference or amplitude difference or a filter extracting a spatial higher order mode, and processed. The processed filter (monitoring) signal is input to the stability monitoring device 26 and the input / output device 20 which are on-line connected stability monitoring means, and the input / output device 20 outputs a filtered LPRM signal to the operator. To be notified.
[0046]
The stability monitoring device 26 monitors the reduction ratio, the resonance frequency (vibration period), the amplitude, and the like representing the stability from the filtered LPRM signal, detects the vibration phenomenon, and discriminates and evaluates the vibration mode. The stability of the state is estimated, and the stability estimation signal is sent to the display device 22 and displayed.
[0047]
In this reactor power monitoring device, when calculating the APRM from the LPRM signal from the neutron detector 15, in addition to the APRM signal obtained by averaging the conventional analog LPRM signal, a spatially dependent vibration mode is selected. A new output fluctuation signal (this output signal is called an RPRM signal or an in-furnace division output region detector signal) is obtained by using a filter to extract the stability, and based on this RPRM signal, Monitor stability. However, as a method for obtaining a filter, a simple one can be used depending on the number and arrangement of LPRM signals to be used, and a simple method for obtaining a filter is included as an option.
[0048]
The output signals obtained through these filters are input to the stability monitoring device 26, and the stability monitoring device 26 sequentially reduces the reduction ratio indicating the stability state of the core 12, the vibration period, and the amplitude indicating the magnitude of fluctuation. The core stability is evaluated online based on these values. From the conventional APRM signal, the core stability related to the mode that changes integrally in the core is monitored. At the same time, the region stability is monitored from the new signal, and the space-dependent power fluctuation in the core is always monitored in real time. Monitor.
[0049]
In addition, the stability monitoring device 26 can estimate a center line of vibration from a phase diagram (Lissajous diagram) of a signal using each filter by preparing a plurality of weight coefficients used for the filter. Alternatively, the center line of vibration can be similarly estimated from the response of the RPRM signal by rotating one type of filter coefficient. That is, it is dominant in the current core by estimating the centerline of vibration from the reduction ratio and amplitude of the APRM signal and the RPRM signal, between the plurality of RPRM signals, or from the RPRM signal obtained by rotating the weighting factor. Estimate spatial higher order modes and monitor core 12 and region stability online.
[0050]
Further, the process control computer 19 periodically reads the core state data and the LPRN signal generated by the core current data measuring device 17 and the A / D converter 18, and according to the read core state data and the LPRM signal, Core state information data (core management data) representing the state in the core 12 at the read timing is generated. For example, the core state information data includes the power distribution in the core 12, the flow rate distribution of the coolant, and other information. The process control computer 19 further stores the generated core state information data in an internal storage area according to the read timing.
[0051]
The stability prediction device 21 predicts the stability of an arbitrary core state based on the latest core state information data stored in the process control computer 19 using a core stability analysis model composed of a physical model. It has become.
[0052]
Here, the outline of the power monitoring flow in the reactor power monitoring apparatus is shown in FIG. 4, and its functional block diagram is shown in FIG.
[0053]
According to FIGS. 4 and 5, each LPRM signal and each core operating state signal are digitized by the A / D converter 18 and converted into LPRM data and core state data.
[0054]
The digitized LPRM data and core state data are output to the process control computer 19 and the monitoring signal calculation device 25 (see steps S18a and 18b in FIG. 4).
[0055]
Similarly, the digital data converted from the APRM signal by the A / D converter 18 is output as APRM data to the process control computer 19 and the monitoring signal calculation device 25.
[0056]
As shown in FIG. 5, the monitoring signal calculation device 25 calculates a monitoring signal (RPRM signal) based on the sent LPRM signal, and transmits the calculated RPRM signal to the stability monitoring device 26. The stability monitoring device 26 receives the transmitted RPRM signal and sequentially evaluates the received RPRM signal to monitor the stability of the core 12.
[0057]
In addition, the stability prediction device 21 periodically fetches the core state information data stored in the process control computer 19, and cooperates with the process control computer 19 in response to an operator's request or periodically. The state and stability of the furnace 10 are predicted (see step S19a).
[0058]
The functions of the process control computer 19, the stability predicting device 21, the monitoring signal calculating device 25, and the stability monitoring device 26 described above are the function modules 19a, 21a to 21d, 25a to 25b, and 26a to 26b in FIG. It is shown schematically.
[0059]
FIG. 4 is a schematic flowchart for explaining details of each functional module shown in FIG.
[0060]
The monitoring signal calculation device 25 applies weight coefficients (filters, filter coefficients) that are preset and held to LPRM data that is sequentially input via the A / D converter 18. That is, the monitoring signal calculation device 25 uses the first weighting factor (first filter) corresponding to the fundamental mode of neutron flux distribution and the second weighting factor (corresponding to the higher order mode of neutron flux distribution) to monitor the LPRM data. The stability monitoring signal (RPRM signal, RPRM data) is calculated by filtering using the second filter (see steps S25a to 25d in FIG. 4).
[0061]
Subsequently, the stability monitoring device 26 determines the parameters (reduction ratio, natural frequency, amplitude, spectrum, etc.) relating to the core stability, channel stability, and region (higher order mode) stability based on the calculated RPRM signal. Are sequentially estimated and evaluated (see steps S26a and 26b in FIG. 4). Regarding higher order mode (region) stability, the higher order mode center line that maximizes the sensitivity of the RPRM signal is estimated to obtain an optimum weighting factor (see steps S26c and 26d), and the RPRM signal is recalculated (see FIG. Step S26e), parameters such as a reduction ratio, natural frequency, amplitude, spectrum, etc. relating to higher-order mode stability are obtained (see Step S26f).
[0062]
Then, the stability monitoring device 26 detects a vibration phenomenon in the core 12 based on the above parameters, and determines a vibration mode in the core 12 (see step S26g).
[0063]
On the other hand, the stability predicting device 21 has a stability analysis model composed of a physical model, and the latest core state information data (output distribution and flow rate) stored in the process control computer 19 in response to a request from an operator. Etc.) from the process control computer 19 (see step S19a), and the core state (in-core state) such as the thermo-hydraulic condition in the core 12 at the operator's request point (arbitrary operating point) is estimated ( (See step S21a in FIG. 4).
[0064]
Then, the stability predicting device 21 performs stability analysis at the operator's request point using the thermal hydraulic condition and the stability analysis model (see step S21b), and the core 12 at the operator's request point. The reduction ratio of the core / region stability is predicted (see step S21c).
[0065]
The stability predicting device 21 also includes the prediction results including the predicted core / regional stability, the operating state including the thermo-hydraulic conditions at the time of prediction, and the actual core / core of the actual core 12 obtained by the stability monitoring device 26. The region stability is accumulated in a database provided in advance in the prediction device 21 (see step S21d).
[0066]
The stability prediction device 21 compares the prediction result with the actual core stability recorded in the database (see step S21e), and executes the learning function. That is, the stability predicting device 21 obtains a deviation between the prediction result (predicted core / region stability) and the monitored core / region stability as a learning function, and corrects the input parameter of the stability model. This eliminates the deviation (see step S21f).
[0067]
Therefore, the reactor power monitoring apparatus of the present invention can detect power fluctuations that may be difficult to detect with a conventional APRM signal, and can contribute to improvement in safety and operating rate.
[0068]
Next, a calculation process for obtaining a desired filter by the monitoring signal calculation device 25 will be described.
[0069]
There are two types of filters that are required. One is a filter obtained by computing each measured LPRM signal that is constantly input from the data sampler 18, and the other is calculated by the higher-order mode calculator 24. This is a filter obtained by extracting the spatial higher order mode distribution.
[0070]
The former is a filter calculated based on the measured signal of each LPRM signal detected sequentially or continuously, and the latter is calculated based on information from the process control computer 19, and changes in operating conditions, for example, control rods. The filter reflects the power distribution in the core 12 as the operating point is changed by operation or recirculation flow rate operation. The latter filter is obtained via the process control computer 19 at the planned change of the reactor core state.
[0071]
The former filter obtained by the monitor signal calculation device 25 is a statistic of each LPRM signal that is an actual measurement signal, that is, a cross-correlation function and a variance value of each LPRM signal are sequentially or every short time interval (several tens of seconds) (continuous). It may be the target.) On the other hand, the higher order mode calculation device 24 of the process control computer 19 calculates the spatial higher order mode neutron flux distribution. If the space-dependent time-varying neutron flux φ (r, t) in the core 12 of the nuclear reactor 10 is configured by superposition of spatial higher-order mode neutron flux distributions,
[Expression 1]

Where φ _m (R) shows the spatial higher-order mode neutron flux distribution in the steady state, n _m (t) represents the amplitude function with which the m-order mode neutron flux fluctuates with time, and satisfies the following eigenvalue equation.
[0072]
[Expression 2]

L (r) and M (r) are a time-independent annihilation (diffusion, absorption, etc.) operator and generation (fission) operator, respectively. _m Represents an eigenvalue.
[0073]
In addition, each eigenvalue K represented by the eigenvalue equation _m The following relationship is established between them.
[0074]
[Equation 3]

That is, from equation (3), the eigenvalue λ of the lowest order mode ₀ Is the largest, and in steady state, the eigenvalue λ of the lowest order mode ₀ Only the last mode remains, so the lowest order mode is referred to as the fundamental mode.
[0075]
The temporal variation of the fundamental mode is expressed by a well-known dynamic characteristic equation. The temporal change of the m-order mode neutron flux (hereinafter, the mode of m> 0 is referred to as a higher-order mode) is expressed by the following dynamic characteristic equations (4) and (5).
[0076]
[Expression 4]

[Equation 5]

In the equation (4), Δρm is a subcriticality of the fundamental mode of the m-th mode and is defined by the following equation.
[0077]
[Formula 6]

[0078]
On the other hand, the following orthogonal relationship is established between the higher-order modes.
[0079]
[Expression 7]

[0080]
However, in a large core such as a nuclear reactor, dynamic higher-order mode distribution distortion is negligible, and since core management is generally performed with a 1/4 core symmetry, there is almost no direct orthogonal relationship between higher-order modes. Satisfied.
[0081]
[Equation 8]

[0082]
In the equations (7) and (8), φ _{m +} (R) means the associated neutron flux, but since we consider a group of neutron flux below, φ _{m +} (R) is the neutron flux distribution φ _m It matches (r). That is, it is expressed by the following formula.
[0083]
[Equation 9]

[0084]
In order to extract only a specific mode component, for example, the temporal change of the n-th mode, from the formulas (1) and (9), the entire neutron flux is multiplied by the n-th mode neutron flux distribution over the entire core. Do it. That is,
[Expression 10]

[0085]
Therefore, the neutron flux mode distribution of interest as a filter may be used. However, what is measured is not the neutron flux distribution Φ (r, t), but LPRM signals that exist in large numbers (for example, 172 locations in a typical 1.1 million KWe class BWR) are used. That is, by using the LPRM signal at a certain location in the core as LPRM (r, t) and using the higher-order mode distribution around it,
[Expression 11]

[0086]
In the equation (11), the APRM signal uses a constant value that does not depend on the place as the weight. This APRM signal includes temporal changes in all neutron flux modes. As described above, it is considered that the transient neutron flux fluctuations in the actual core generally include many higher-order modes. It has been confirmed analytically and from the evaluation result of the actual machine vibration data that the fluctuation component of the direction primary mode is dominant. Conventionally, the core stability was evaluated by the APRM signal, but more accurately, it is necessary to use a signal weighted in the basic mode.
[0087]
Based on such a principle, it is optimal to use a higher-order mode distribution as a filter coefficient (see US Pat. No. 5,406,598), but at present, the higher-order mode distribution is obtained one by one and used as a filter coefficient. I can't do it.
[0088]
For the above reason, this embodiment of the present invention brings about an effect equal to or higher than the method using the higher-order mode distribution by devising and using a simpler method, and further, the reactor output that is easier to implement. A monitoring device is proposed.
[0089]
First, when the higher-order mode distribution in the actual core state cannot be obtained, the neutron flux distribution φ of the bare homogeneous cylindrical core _lmn Using (r, θ, z) and assuming that the index in three directions is (l, m, n),
[Expression 12]

[0090]
In the basic mode, which is a mode related to the reactor core integrated vibration, m = 0 and l = n = 1. Further, in the circumferential primary mode that is a mode related to the region vibration, it corresponds to l = m = n = 1.
[0091]
Therefore, for the z (axis) direction mode, the basic mode and the circumferential primary mode coincide. In the higher-order mode m ≧ 1, since it is periodic with respect to the azimuth angle θ, m linear zeros are added. Therefore, in the circumferential primary mode (m = 1), two different distributions that are orthogonal to each other exist for the same eigenvalue (that is, the eigenvalue is degenerate double).
[0092]
Since we are paying attention to the stability of the neutron flux distribution in the cross-sectional direction of the core, the filter coefficient is m-th order Bessel function J _m (X) is used. Since the basic mode corresponds to m = 0 and n = 1, the filter coefficient is J ₀ (X _0,1 If r / R) is used, m = n = 1 in the circumferential primary mode of the mode related to the core stability. ₁ (X _1,1 By using r / R) cos θ, it is possible to selectively extract the modes related to the region stability.
[0093]
Since the m-th order Bessel function is a mode distribution in a bare homogeneous core system, it does not coincide with the mode distribution of the actual core, but the core management is performed in a ¼ core symmetry. For this reason, the neutron flux distribution shows a distribution that is relatively close to the m-th order Bessel function for the core cross section (see FIG. 6), so that it is considered that sufficient accuracy can be obtained for the purpose of stability monitoring.
[0094]
The right figure in FIG. 6 shows the actual circumferential primary mode distribution in a typical BWR actual core. In contrast to the right diagram in FIG. 6, the left diagram shows the distribution of the primary Bessel function corresponding to the circumferential primary mode distribution in the homogeneous furnace body. It can be seen that the Bessel function distribution approximates the circumferential primary mode distribution in the actual core.
[0095]
However, since the circumferential primary mode is a multiple root, J ₁ (X _1,1 r / R) cosθ and J orthogonal to it ₁ (X _1,1 There are two types: r / R) sinθ. Therefore, if the actual vibration mode is a distribution orthogonal to the prepared filter distribution, the vibration mode cannot be extracted by the filter. Since the region vibration mode actually generated cannot be predicted accurately, two circumferential primary modes orthogonal to each other as a filter, that is, J ₁ (X _1,1 r / R) cosθ and J ₁ (X _1,1 Two types of r / R) sinθ may be prepared in advance.
[0096]
Here, for convenience, the stability monitoring signal obtained by using the fundamental mode distribution as a weight is RPRM0, and the stability monitoring signal obtained by using the circumferential primary mode weights orthogonal to each other is RPRM1. / RPRM2 will be referred to. FIG. 7 shows various RPRM signal generation procedures using such a filter.
[0097]
FIG. 7 shows a conceptual diagram of a weight distribution when using an APRM signal (averaged signal). All currently used APRM signals are obtained using the same weighting factor, but the stability monitoring signal (RPRM signal) in the present invention uses a weight distribution corresponding to a higher-order mode to be monitored. It shows that you want to. That is, the stability monitoring signal RPRM0 corresponding to the basic mode uses the basic mode as a weight, and the stability monitoring signal RPRM1 / RPRM2 corresponding to the circumferential primary mode uses the circumferential primary mode distribution orthogonal to each other as the weight. .
[0098]
In particular, at the time of region vibration in which a phase difference occurs spatially, since there is a phase difference between the LPRM signals, the conventional APRM signals cancel each other, resulting in a decrease in signal sensitivity. On the other hand, in the RPRM1 / RPRM2 signal, the phase difference is corrected by the weighting factor used for the filter, canceling does not occur, and a highly sensitive RPRM signal is obtained. These circumstances are shown in FIG.
[0099]
In FIG. 8, in the conventional method, since the APRM signal is obtained by using the same weighting factor for all LPRM signals, the obtained APRM signals cancel each other out, and the APRM signal itself Will be attenuated. On the other hand, in the RPRM signal according to the present invention, since cancellation of individual LPRM signals does not occur, it is shown that a high-order mode response of interest can be extracted.
[0100]
Next, an example of applying vibration monitoring using a filter to simulated vibration data using a three-dimensional dynamic characteristic analysis code will be described.
[0101]
First, FIG. 9 shows an example applied to core-integrated vibration. Although the vibration phenomenon is accurately captured by the fundamental mode stability monitoring signal RPRM0, the circumferential primary mode stability monitoring signal RPRM2 (same for RPRM1) contains almost no vibration component. That is, it can be seen from the difference in response between the stability monitoring signal RPRM0 and the stability monitoring signal RPRM1 / RPRM2 that this vibration phenomenon is the reactor core integrated vibration.
[0102]
Similarly, an example applied to region vibration is shown in FIG. Compared with RPRM0 / RPRM1, the vibration of RPRM2 is significantly larger and the period is clear. From this, it can be determined that the vibration is a region vibration in a direction (in this example, the vibration center line is 135 degrees) that matches the distribution of the weighting factor of RPRM2. FIG. 11 shows the relationship between the weighting factor distribution of the RPRM signal and the RPRM signal response.
[0103]
Since the analysis result used in FIG. 10 is used in FIG. 11, RPRM2 uses a weighting factor that matches the vibration direction of the region vibration, and RPRM1 uses a weighting factor that is orthogonal thereto. Here, if an RPRM signal obtained by using a weighting coefficient intermediate between them, that is, a weight distribution shifted by 45 degrees from RPRM1 / RPRM2, is named RPRM3, the amplitude of RPRM3 becomes an intermediate value between the two. I understand that.
[0104]
That is, the direction of the region vibration (the center line of the vibration, that is, the direction of the zero point of the spatial higher-order mode appearing as the region vibration) is closely related to the amplitude (sensitivity) of the RPRM signal through the distribution of the weight coefficient. It is shown that. J as the weighting factor ₁ (X _1,1 The change in the average amplitude of the RPRM signal when r / R) sin θ is used and the center line angle where the sign (positive or negative) of the weighting coefficient is changed, that is, the azimuth angle θ is changed from 0 degrees to 180 degrees at intervals of 5 degrees. Is shown in FIG.
[0105]
When the center line (135 degrees) of vibration and the azimuth angle θ coincide, the amplitude of the RPRM signal is maximum, that is, the sensitivity is maximum, and conversely, the sensitivity is minimum at an angle (45 degrees) orthogonal thereto. Therefore, two kinds of weighting coefficients (J ₁ (X _1,1 r / R) cosθ and J ₁ (X _1,1 If r / R) sin θ) is prepared, the amplitude of at least one of the RPRM signals is in the shaded area of FIG. Therefore, in order to monitor the region vibration, it is necessary and sufficient to prepare two kinds of weighting factors similar to the circumferential first-order mode distribution orthogonal to each other.
[0106]
Also, as can be seen from FIG. 12, the distribution angle of the weighting factor is made variable (or a large number of weighting factors are prepared by changing the angle little by little as shown in this figure to obtain each RPRM signal). The direction of the region vibration can be determined from the angle at which the amplitude of the RPRM signal is maximized. However, it is possible to estimate the direction of the region vibration from the behavior of the phase diagram of the RPRM1 signal and the RPRM2 signal using two kinds of weighting factors orthogonal to each other. The situation is shown in FIG.
[0107]
In FIG. 13, the change of the response of the RPRM1 signal and the RPRM2 signal using two kinds of weighting factors orthogonal to each other with respect to the direction of the centerline of the weighting factor is shown in a phase diagram of both signals. From the azimuth angle θ at which the response of the phase diagram is maximized, the direction of the center line of vibration is known. That is, when RPRM1 is sufficiently larger than RPRM2 (lower right direction in the figure), the higher-order mode instability corresponding to RPRM1 is dominant, so the vibration direction matches the higher-order mode distribution. Conversely, in the upper left direction of FIG. 13, the direction of vibration is a direction orthogonal to the higher-order mode distribution. If these directions are intermediate (near the diagonal line), the vibration direction is in the middle of the center lines of both distributions.
[0108]
That is, if the amplitude of any RPRM signal of RPRM1 / RPRM2 is significantly biased, the direction of the region vibration is closer to the direction of the center line of the weight coefficient distribution of the RPRM signal having a large amplitude, and if the amplitude is the same, both It can be estimated that the direction is intermediate. Therefore, constantly monitoring the phase diagrams of the RPRM1 signal and the RPRM2 signal and displaying them on the display device 22 is also an effective means for monitoring the region stability.
[0109]
Here, when both the RPRM signals have the same amplitude, the determination of whether the vibration is the region vibration or the core-integrated vibration may be made by comparing with the amplitude of the RPRM0 signal as shown in FIG. That is, if the amplitude of the RPRM0 signal is significantly larger than the amplitude of the RPRM1 / RPRM2 signal, it can be determined that the vibration is the reactor core integrated vibration, and conversely, if it is smaller, it is the region vibration.
[0110]
Accordingly, not only the reduction ratio, which is an index of stability obtained from each RPRM signal, but also the mutual relationship between the amplitude, period, and phase space of each signal is necessary for monitoring the output vibration phenomenon. Relates to a means for providing a method for judging a vibration phenomenon based on the comprehensive judgment (FIGS. 4 and 5).
[0111]
Furthermore, the above-described method can be applied to the monitoring of reactor power stability including higher modes. That is, the following values are used as weighting factors.
[0112]
[Formula 13]

The symbols used here are the same as in equation (12).
[0113]
Using the weighting factor extended to this arbitrary higher-order mode, the azimuth angle θ is rotated every certain time, and the azimuth angle with the largest output amplitude is set. This azimuth angle differs depending on the higher-order mode to be used, that is, for each value of (m, n) in equation (13). The RPRM signal for each higher-order mode is obtained using the weighting coefficient obtained by substituting the set optimum azimuth angle into the equation (13). Here, since the azimuth angle is optimized, only one type of RPRM signal is required.
[0114]
Since the obtained RPRM signal has temporal information of each higher-order mode, stability monitoring parameters such as a reduction ratio, amplitude, and period are extracted from this signal. FIG. 14 shows an example of monitoring.
[0115]
Based on the method described above for any higher-order mode from the basic mode, the direction of each mode (upper right in the figure), the reduction ratio and period (frequency) of the RPRM signal representing the time response of each mode (lower right in the figure), The response waveform of each RPRM signal and the transition of the amplitude are displayed for each mode.
[0116]
In this example, the fundamental mode component, that is, the core-integrated vibration component is the main component. For example, only the first-order mode component is included, but other higher-order mode components are also included. It turns out that the higher-order mode component itself is higher.
[0117]
Another monitoring item is the spectrum of each RPRM signal. FIG. 15 shows an example in which the response waveform, amplitude transition, and spectrum (self power spectral density: APSD) of each RPRM signal are displayed.
[0118]
The stability of each mode can be determined from the peak height and the sharpness of the APSD. Further, whether or not a peak, that is, a harmonic peak, appears at an integer multiple of the fundamental frequency of the spectrum. Thus, it can be determined whether or not the mode component is dominantly developed as vibration.
[0119]
For example, FIG. 16 shows the APSD (self-power spectral density) of the RPRM signal at the time of the developed output vibration, and the harmonics having a frequency twice or three times the fundamental frequency (about 0.35 Hz) clearly appear. However, they do not appear in the spectrum of FIG. 15 (where the peaks appearing above 1 Hz are for other reasons not related to stability).
[0120]
Therefore, in the example shown here, a relatively high reduction ratio is shown and higher-order mode components are included, but the amplitude is at most 3% even at the peak-to-peak value, and the stability deteriorates. Since the non-linear harmonic component that becomes noticeable at the time of oscillation is not included, it can be seen that the entire core is relatively stable and is not in a state particularly related to safety. The fundamental mode component having the largest amplitude and the Nth-order mode component whose stability itself is deteriorated may be monitored intensively.
[0121]
Here, monitoring is performed using all LPRM signals that can be used in the reactor core. As a display method, these change components are displayed in three dimensions in real time, so that spatial fluctuations can be visually observed. It is possible to grasp.
[0122]
[Outside 1]

[0123]
[Expression 14]

The standardized deviation value is obtained, and those fluctuations are displayed in real time on a three-dimensional monitoring device screen corresponding to the arrangement of the detectors in the core. FIG. 17 shows an example of the display.
[0124]
In the display example of FIG. 17, a plot of about 2 seconds for one vibration period is shown, but actually, such a three-dimensional view is displayed on the display device 22 as an animation in almost real time. Further, a contour map of each LPRM signal amplitude is displayed at the bottom of FIG. 17, but if the color to be displayed is changed according to the amplitude value, the spatial behavior of the output fluctuation in the reactor core is displayed. 19, it becomes easy to grasp visually.
[0125]
Also, instead of the LPRM signal, if the LPRM signal for the so-called higher-order mode in which each LPRM signal is modified with the weight optimized by the equation (13), the fluctuation of the higher-order mode component used in the weight is used as an animation. It is possible to display as well.
[0126]
The above is a method for monitoring all or most of the LPRMs arranged uniformly in the core. However, a method for more simply monitoring the region stability using a smaller number of limited LPRM signals. This will be described below.
[0127]
For example, as shown in FIG. 6, in the circumferential primary mode, the core periphery has a high distribution. Therefore, the contribution to the LPRM signal of the core center of the high-order mode in terms of region stability compared to the peripheral portion. It is considered small. Therefore, if the core stability is monitored by the APRM signal as in the conventional case, only the LPRM signal in the peripheral portion needs to be used for the region stability monitoring. Furthermore, if the LPRM signal of the LPRM detector that is substantially equidistant from the core center is used, the same value can be approximately used as the weighting factor.
[0128]
FIG. 18 shows an example of arrangement of LPRM signals in the core used for simple core stability monitoring using fewer limited LPRM signals. In this example, five LPRM signals in the center of the core are used for monitoring the core stability, and ten LPRM signals in the periphery of the core are used for monitoring the regional stability, for a total of 15 LPRM signals. 0.2 is uniformly used as the signal weighting coefficient for monitoring the core stability, and ± 0.2 is used as the weighting signal for monitoring the region stability.
[0129]
For example, in the region stability monitoring LPRM signal in FIG. 18, a signal using a weight of 0.2 for (3-7) LPRM and a weight of −0.2 for LPRM (8-2) is RPRM1, A signal in which 0.2 is used for the LPRM of (1-5) and a weight of -0.2 is used for the LPRM of (6-10) is RPRM2. Further, it is assumed that each LPRM signal is standardized by all DC values. By this normalization, a uniform value is used as a weighting coefficient, and a deviation component from the DC value of each signal is extracted.
[0130]
FIG. 19 shows an example applied to core stability monitoring, and FIG. 20 shows an example applied to region stability monitoring. Since the amplitude of RPRM0 is dominant at the time of core-integrated vibration, and either RPRM1 or RPRM2 is dominant at the time of region vibration, stability monitoring is performed as in the case of using the LPRM signal. However, since the influence from other modes cannot be deleted safely, it is difficult to monitor only by the reduction ratio, and an evaluation in combination with the amplitude is required.
[0131]
19 and 20 show a simulation of a vibration phenomenon using a three-dimensional stability analysis code built in the process control computer 19 in the stability predicting device 21, and a nuclear instrumentation signal (LPRM) at that time is proposed in the present invention. RPRM0 (core stability monitoring signal corresponding to the basic mode) signal and RPRM1 (regional stability monitoring corresponding to the circumferential primary mode) from the monitoring signal calculation device 25 and the stability monitoring device 26, which are obtained using the weights obtained. Signal) shows the response of the signal.
[0132]
The above is the contents related to the sequential stability monitoring method from the time series data. The stability prediction function based on the physical model by the stability prediction device 21 will be further described with reference to FIG.
[0133]
First, the calculation flow in the stability prediction function is shown in FIG. 21 in comparison with the calculation flow of stability monitoring in the stability monitoring device 26.
[0134]
When an operator (user) performs stability prediction at an arbitrary operating point having a predetermined operating condition, the operating condition (output and coolant flow rate, etc.) to be predicted is input to the stability predicting device 21 (FIG. 21 step S30). The stability predicting device 21 accesses the latest core state information data (core management data; see D1 in FIG. 21) stored at the input time stored in the process control computer 19 and receives the power distribution (see step S31). . Then, the stability prediction apparatus 21 first selects the fuel assembly having the highest output from the received output distribution, and sets it as an assembly for obtaining channel stability.
[0135]
Next, the stability prediction apparatus 21 collects several fuel assemblies that are close in terms of thermohydraulic conditions according to the output distribution and divides them into a small number of channel groups for core stability analysis (step). (See S32).
[0136]
Similarly, the stability predicting device 21 determines the fuel assembly based on the vibration distribution center line (see L1) estimated by the output distribution and the stability monitoring function of the stability monitoring device 26, and the higher-order mode weight coefficient. Some of them are collectively divided into a small number of channel groups for region stability analysis (see step S33).
[0137]
In addition, higher-order mode subcriticality is required for region stability analysis, but since higher-order mode is not evaluated in the current core monitoring system, higher-order mode subcriticality is a function of power distribution in advance. It is calculated as G and calculated using the output distribution obtained from the monitoring system (see step S34).
[0138]
The procedure for determining the higher order mode subcriticality is as follows. That is, the root mean square R of the next power distribution is obtained with the power distribution of a certain fuel assembly as Prn.
[0139]
[Expression 15]

[0140]
Similarly, the root mean square RL with respect to the distance from the core center of the fuel assembly is obtained.
[0141]
[Expression 16]

[0142]
Then, the higher-order mode subcriticality is expressed as a function of the difference between the above two indices R and RL. Also, a function H of the operating conditions (output P, coolant flow rate F) is fitted in advance as the subcriticality for the operating conditions to be predicted. Then, the subcriticality Δρ for the operating condition to be predicted is
[Expression 17]

Given in.
[0143]
In addition to this, variables that vary depending on the operating conditions, such as bypass section flow rate, channel axial direction output distribution, etc. are given in advance as a function of operating conditions, and operating conditions given core management data (core state information data), The variable value in the predicted operation state is set from the deviation of the state from the operation condition for which stability is to be predicted. Also, variable values that can be estimated from the operating conditions using the steady heat balance, for example, the core pressure, the core inlet / outlet temperature, and the like are calculated and set from the heat balance physical model (see H1 in FIG. 21) (f).
[0144]
As described above, when the thermal hydraulic conditions including the number of channel divisions in the predicted operation state, the value of subcriticality, the value of the power distribution, and the like are obtained, the stability predicting device 21 determines The values of the power distribution and the like are taken into the physical model for stability analysis (see M1 in FIG. 21) (see step S35), and the core stability, channel stability and region stability are analyzed (see step S36). The channel / core stability reduction ratio and natural frequency are calculated (see step S37), respectively, and the region stability reduction ratio and natural frequency are calculated (see step S38).
[0145]
Further, as shown in FIG. 21, the prediction model obtained by the stability prediction device 21 can always use stability information including core state information data obtained from the plant (ie, the reactor 10). (Refer to step S39) By executing the learning function at the time of executing the prediction process, the stability monitoring process shown in steps S41 to S47 substantially corresponding to steps S25a to 25d and steps S26a to 26g in FIG. The bias between the stability obtained from the plant time series data including the LPRM data and the stability obtained from the prediction function can be corrected (see step S40). As adjustment parameters by learning, power distribution and void reactivity coefficient are considered for core / regional stability, recirculation system transfer characteristics for core stability alone, and higher-order mode subcriticality for region stability.
[0146]
For example, when the prediction function is explicitly activated by the user, and when the state changes to some extent from the current operating point due to an operating condition or a transient event, the stability monitoring function and stability by the stability monitoring device 26 are improved. The learning function is performed with a certain frequency by automatically starting the prediction function by the sex prediction device 21. The void reactivity coefficient is a parameter that affects both core stability and region stability, but the transfer characteristics of the recirculation system are parameters that only affect core stability, and subcriticality is a parameter that affects only region stability. Therefore, first, adjustment learning of both stability is performed by the void reactivity coefficient, and then adjustment learning of core stability is performed by the transfer characteristic of the recirculation system, and region stability adjustment learning is performed by the subcriticality. The above learning function is realized by the stability predicting device 21 of FIG.
[0147]
That is, in the present invention, a new monitoring signal using a weighting factor considering the fundamental mode and higher-order mode distribution of the neutron flux distribution is obtained by the monitoring signal calculation device 25, and the stability is reduced by the stability monitoring device 26 from the signal response. It is possible to determine the ratio, amplitude and phase relationship for each coefficient, and to comprehensively and comprehensively determine the spatial vibration phenomenon in the core from them. In addition, the stability prediction apparatus 21 uses the latest operation management data and can predict the change in core stability when the operation state changes by the stability prediction based on the physical model, thereby enabling more flexible operation. .
[0148]
In the reactor power monitoring apparatus of this embodiment, it is possible to detect and judge more accurately and including output fluctuations that are difficult to detect with the conventional signal, compared to the conventional APRM signal that averages the analog signal as it is, It is possible to improve safety and improve the operation rate.
[0149]
【The invention's effect】
As described above, in the reactor power monitoring apparatus according to the present invention, the fundamental mode distribution of the neutron flux in the reactor core state in the core and Orthogonal to each other By using the weighting coefficient corresponding to the higher-order mode distribution to filter the conventional detection signal to generate the stability monitoring signal, the reactor power fluctuation that is difficult to detect with the conventional APRM signal can be detected as described above. Therefore, it is possible to accurately monitor using the detection signal, and it is possible to improve the safety and the operating rate.
[0150]
While the reactor power monitoring device of the present invention can accurately monitor reactor power fluctuations that are often overlooked by conventional reactor power monitoring methods, it can predict changes in core stability when operating conditions change, and more Flexible operation is possible.
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing an embodiment of a reactor power monitoring apparatus according to the present invention.
2 is a cross-sectional plan view of a core showing an example of arrangement of neutron flux monitors in the core of the reactor shown in FIG. 1;
3 is a plan sectional view showing an example of arrangement of the neutron flux monitor shown in FIG. 2 in the core axis direction (vertical direction).
FIG. 4 is a flowchart schematically showing a flow of processing of stability monitoring and stability prediction in the reactor power monitoring apparatus according to the present invention.
FIG. 5 is a block diagram showing functional modules of a reactor power monitoring apparatus according to the present invention.
FIG. 6 shows an example of power distribution used as a filter weight in the reactor power monitoring apparatus according to the present invention, and is a comparison diagram of the Bessel function distribution and the primary mode distribution in the circumferential direction of the actual machine.
FIG. 7 is a diagram showing a method of creating a new RPRM signal introduced in the reactor power monitoring apparatus according to the present invention.
FIG. 8 is a diagram showing the effect of a filter by the reactor power monitoring apparatus according to the present invention.
FIG. 9 shows an example of output stability monitoring in the reactor power monitoring apparatus according to the present invention, and is a diagram showing an example of response of an RPRM signal indicating an application result at the time of reactor core integrated vibration.
FIG. 10 is a view showing an example of output stability monitoring in the reactor power monitoring apparatus according to the present invention, and showing a response example of an RPRM signal indicating an application result at the time of region vibration.
FIG. 11 is a diagram showing the relationship between the distribution of weighting factors and the RPRM response, which is an example of power stability monitoring in the reactor power monitoring apparatus according to the present invention. The figure which compared the difference in the response of the monitoring signal when rotating in the direction.
FIG. 12 is a diagram showing a relationship between a weighting coefficient center line angle and an RPRM amplitude, which is an example of power stability monitoring in the reactor power monitoring apparatus according to the present invention; The figure which showed the response | compatibility on the phase diagram of a signal, and the response | compatibility with the centerline direction of a region vibration.
FIG. 13 is a view showing an example of output stability monitoring in the reactor power monitoring apparatus according to the present invention, showing the RPRM1 / RPRM2 phase diagram and the direction of region vibration.
FIG. 14 is a diagram showing a high-order mode monitoring example in which each mode component is monitored from the viewpoint of the reduction ratio, natural frequency, monitoring signal raw waveform and amplitude in the reactor power monitoring apparatus according to the present invention.
FIG. 15 is a diagram showing a high-order mode spectrum monitoring example in which each high-order mode component is monitored from the viewpoint of a raw monitor signal waveform, its amplitude, and the spectrum of the monitor signal in the reactor power monitoring apparatus according to the present invention.
FIG. 16 is a diagram showing an example of a harmonic peak appearing in a spectrum during developed vibration in the reactor power monitoring apparatus according to the present invention.
FIG. 17 shows a three-dimensional display example of standardized LPRM signal fluctuations in the reactor power monitoring apparatus according to the present invention. The fluctuations of the monitoring signal, in this case, the smoothed LPRM signal, are shown for each LPRM. The figure which shows the example which displayed the animation sequentially in the position corresponding to arrangement | positioning in the core of a detector.
FIG. 18 is a diagram showing an arrangement example of the LPRM detector for monitoring the core operability in the core.
FIG. 19 is an example of a simplified core stability monitoring in the reactor power monitoring apparatus according to the present invention, and shows a result of an example of the core stability monitoring by the simplified stability monitoring signal, showing the result applied to the integral vibration of the core.
FIG. 20 is a diagram of a region stability monitoring example using a simple stability monitoring signal, showing a result of application to region vibration, which is a simple core stability monitoring example in the reactor power monitoring apparatus according to the present invention.
FIG. 21 is a flowchart showing the flow of calculation in the stability prediction function together with the flow of calculation in the sequential stability monitoring function in the reactor power monitoring apparatus according to the present invention.
[Explanation of symbols]
10 Reactor
11 Reactor pressure vessel
12 core
13 Neutron flux detector (neutron flux measuring means)
14 Fuel assembly
15 Neutron flux detector (LPRM)
16 Control rod
17 Core current data measuring instrument
18 A / D converter
19 Process control computer (neutron flux distribution calculation means)
20 I / O device (I / O means)
21 Stability prediction device (stability prediction means)
22 Display device
24 Higher order mode calculator
25 Monitoring signal calculation device (filter calculation device)
26 Stability monitoring device (stability monitoring means)

Claims (14)

A plurality of neutron flux measuring means arranged in the core of the reactor;
A process control computer that estimates the power state in the core based on the neutron flux detection signal measured by these neutron flux measuring means;
A monitoring signal calculation device comprising a filter for extracting the characteristics of the detection signal fluctuation based on the neutron flux detection signal;
A stability monitoring device for evaluating the stability of the core from the value of the neutron flux detection signal subjected to the required filter in this monitoring signal calculation device;
A display device for displaying the evaluation result;
Using the function of the process control computer, comprising a stability prediction device that predicts an arbitrary operating state ,
The monitoring signal calculation device generates a monitoring signal by filtering the neutron flux detection signal with a weighting factor corresponding to a fundamental mode distribution of a neutron flux in a reactor core state and a higher-order mode distribution orthogonal to each other ,
The reactor power monitoring device, wherein the stability monitoring device is configured to estimate and evaluate the stability of the core based on a monitoring signal obtained by the filtering process. The stability monitoring device evaluates two types of output signals from the monitoring signal calculation device on the phase plane when calculating weighting coefficients corresponding to higher-order mode distributions orthogonal to each other , The reactor power monitoring apparatus according to claim 1 , wherein the reactor power monitoring apparatus is configured to estimate a vibration center line of region stability. When calculating the weighting coefficient corresponding to the higher-order mode distribution orthogonal to each other , the monitoring signal calculation device rotates an arbitrary higher-order mode distribution weight in the core azimuth direction with the obtained filter, and the response The reactor power monitoring apparatus according to claim 1, wherein the stability monitoring apparatus estimates the center line of the higher-order mode distribution from the amplitude. When calculating weighting factors corresponding to higher-order mode distributions orthogonal to each other , the reduction ratio is reduced by the stability monitoring device from the output signal from the monitoring signal calculation device obtained by applying the optimum weighting factor in the core azimuth direction. The reactor power monitoring apparatus according to claim 1, wherein the reactor power monitoring apparatus is configured to obtain an oscillation period and an amplitude and to evaluate arbitrary higher-order mode stability. While obtaining the spectral density as stability evaluation method with different indicators to the monitoring signal computing device from the value of the neutron flux detection signals required filter is applied in the core stability monitoring device decay ratio, the stability monitoring 5. The apparatus according to claim 1, wherein the apparatus is configured to evaluate the importance of the contribution to the stability of the higher order mode by confirming the presence or absence of a harmonic component corresponding to an integral multiple of the fundamental frequency in the obtained spectral density. Reactor power monitoring device. Said monitoring signal computing device while outputting an output signal related to the fundamental mode and higher order modes obtained by, the decay ratio obtained was evaluated for stability monitoring device the output signal, the oscillation period, outputs an amplitude and spectral 2. The reactor power monitoring apparatus according to claim 1, wherein the reactor power monitoring apparatus is configured to display on a screen together with a signal response. The monitoring signal calculation device obtains a smoothed variation of the neutron flux detection signal from the neutron flux measuring means arranged in the core, and the stability monitoring device inputs this variation, and sequentially 2. The reactor power monitoring apparatus according to claim 1, wherein a spatial fluctuation of the output in the core is evaluated by displaying an animation on a three-dimensional coordinate corresponding to the measurement position by a display device. The monitoring signal calculation device multiplies the smoothed variation of the neutron flux detection signal from the neutron flux measuring means arranged in the core by a weight distribution corresponding to each higher order mode, and monitors the higher order mode fluctuation stability. by animating sequentially in the display device in the three-dimensional coordinates to correspond to the core of the measuring position by a device, according to claim 1, characterized in that so as to evaluate the spatial variation of the higher-order modes in the core The reactor power monitoring device described . When the monitoring signal calculation device calculates a weighting factor corresponding to a high-order mode distribution orthogonal to each other and the neutron flux detection signal that can be used is small, the neutron flux detection of the peripheral portion that is equidistant from the core center portion calculating a simple output signal using the signal, stability monitoring device from the calculated output signal obtains the decay ratio and amplitude, claim, characterized by being configured to evaluate the regional stability 1 The reactor power monitoring device described . When calculating the weighting coefficient corresponding to the higher-order mode distribution orthogonal to each other , the monitoring signal calculation apparatus normalizes each neutron flux detection signal with each DC value, and the absolute values are equal, and the sign is positive or negative with half each. 10. A simple output signal is calculated by using the weighting factor, and the stability monitoring device obtains a reduction ratio and an amplitude from the output signal, and evaluates the region stability. The reactor power monitoring device described. The supervisory signal calculation device calculates an output signal obtained by using two kinds of weighting coefficients whose center lines for inverting the signs of the weighting coefficients are orthogonal to each other when calculating the weighting coefficients corresponding to the higher-order mode distributions orthogonal to each other. The reactor power monitoring device according to claim 1 or 9, wherein the stability monitoring device is configured to easily evaluate the region stability by obtaining the reduction ratio and the amplitude from the output signal. The process control computer or stability prediction device is equipped with a stability analysis model based on the physical model based on the latest core management data. The reactor power monitoring apparatus according to claim 1, wherein the reactor power monitoring apparatus is configured to predict. The stability predicting device is a parameter that changes as the operating point is changed, and a function that is not physically modeled is approximated by a function within the operating point range that changes in advance, as a deviation from the latest operation management data. The reactor power monitoring apparatus according to claim 1 or 13 , wherein the reactor power monitoring apparatus is configured to perform parameter evaluation at an operating point for predicting stability. By comparing the core / region stability reduction ratio predicted by the stability prediction device with the core / region stability reduction ratio during operation evaluated by the monitoring signal calculation device and the stability monitoring device. The reactor power monitoring apparatus according to claim 1 , wherein the stability prediction apparatus improves the accuracy of the prediction function by learning.

Images