#pragma GCC optimize("Ofast", "unroll-loops")

#include <bits/stdc++.h>

using namespace std;

using ll = long long;

int N, P;
ll K;
vector<int> A, B;

void input(void){
    cin >> N >> K >> P;
    A.resize(N);
    for (int& ai : A)
        cin >> ai;
    B.resize(N);
    for (int& bi : B)
        cin >> bi;
    sort(A.begin(), A.end());
    sort(B.begin(), B.end());
}

int _count_B(int l, int r){
    if (r <= l) return 0;
    auto l_iter = lower_bound(B.begin(), B.end(), l);
    auto r_iter = lower_bound(B.begin(), B.end(), r);
    return (r_iter - l_iter);
}

int count_B(int ai, int v){
    return (_count_B(0, v - ai + 1) + _count_B(P - ai, P + v - ai + 1));
}

ll count_v(int v){
    ll ret = 0;
    for (int ai : A)
        ret += count_B(ai, v);
    return ret;
}

int solve(void){
    int l = -1, r = 2 * P - 2;
    while (r - l > 1){
        int mid = (l + r) / 2;
        if (count_v(mid) < K) l = mid;
        else r = mid;
    }
    return r;
}

int main(void){
    input();
    cout << solve() << endl;
    return 0;
}