// code template is in https://github.com/drken1215/algorithm/blob/master/template_minimum.cpp
#pragma GCC optimize("Ofast")
#pragma GCC optimize("unroll-loops")

#include <bits/stdc++.h>
using namespace std;


//------------------------------//
// Utility
//------------------------------//

using ll = long long;
using i128 = __int128_t;
using u128 = __uint128_t;
using pint = pair<int, int>;
using pll = pair<long long, long long>;
using tll = array<long long, 3>;
using fll = array<long long, 4>;
using vint = vector<int>;
using vll = vector<long long>;
using dint = deque<int>;
using dll = deque<long long>;
using vvint = vector<vector<int>>;
using vvll = vector<vector<long long>>;
using vpll = vector<pair<long long, long long>>;
template<class T> using min_priority_queue = priority_queue<T, vector<T>, greater<T>>;

template<class S, class T> inline bool chmax(S &a, T b) { return (a < b ? a = b, 1 : 0); }
template<class S, class T> inline bool chmin(S &a, T b) { return (a > b ? a = b, 1 : 0); }
template<class S, class T> inline auto maxll(S a, T b) { return max(ll(a), ll(b)); }
template<class S, class T> inline auto minll(S a, T b) { return min(ll(a), ll(b)); }
template<class T> auto max(const T &a) { return *max_element(a.begin(), a.end()); }
template<class T> auto min(const T &a) { return *min_element(a.begin(), a.end()); }
template<class T> auto argmax(const T &a) { return max_element(a.begin(), a.end()) - a.begin(); }
template<class T> auto argmin(const T &a) { return min_element(a.begin(), a.end()) - a.begin(); }
template<class T> auto accum(const vector<T> &a) { return accumulate(a.begin(), a.end(), T()); }
template<class T> auto accum(const deque<T> &a) { return accumulate(a.begin(), a.end(), T()); }

#define REP(i, a) for (long long i = 0; i < (long long)(a); i++)
#define REP2(i, a, b) for (long long i = a; i < (long long)(b); i++)
#define RREP(i, a) for (long long i = (a)-1; i >= (long long)(0); --i)
#define RREP2(i, a, b) for (long long i = (b)-1; i >= (long long)(a); --i)
#define EB emplace_back
#define PF push_front
#define PB push_back
#define MP make_pair
#define FI first
#define SE second
#define ALL(x) x.begin(), x.end()
#define COUT(x) cout << #x << " = " << (x) << " (L" << __LINE__ << ")" << endl

// input
template<class T> istream& operator >> (istream &is, vector<T> &P)
{ for (int i = 0; i < (int)P.size(); ++i) cin >> P[i]; return is; }
template<class T> istream& operator >> (istream &is, deque<T> &P)
{ for (int i = 0; i < (int)P.size(); ++i) cin >> P[i]; return is; }
template<class T> istream& operator >> (istream &is, vector<vector<T>> &P)
{ for (int i = 0; i < (int)P.size(); ++i) cin >> P[i]; return is; }

// output
template<class S, class T> ostream& operator << (ostream &s, const pair<S, T> &P)
{ return s << '<' << P.first << ", " << P.second << '>'; }
template<class T> ostream& operator << (ostream &s, const array<T, 2> &P)
{ return s << '<' << P[0] << "," << P[1] << '>'; }
template<class T> ostream& operator << (ostream &s, const array<T, 3> &P)
{ return s << '<' << P[0] << "," << P[1] << "," << P[2] << '>'; }
template<class T> ostream& operator << (ostream &s, const array<T, 4> &P)
{ return s << '<' << P[0] << "," << P[1] << "," << P[2] << "," << P[3] << '>'; }
template<class T> ostream& operator << (ostream &s, const vector<T> &P)
{ for (int i = 0; i < P.size(); ++i) { if (i > 0) { s << " "; } s << P[i]; } return s; }
template<class T> ostream& operator << (ostream &s, const deque<T> &P)
{ for (int i = 0; i < P.size(); ++i) { if (i > 0) { s << " "; } s << P[i]; } return s; }
template<class T> ostream& operator << (ostream &s, const vector<vector<T>> &P)
{ for (int i = 0; i < P.size(); ++i) { s << endl << P[i]; } return s << endl; }
template<class T> ostream& operator << (ostream &s, const set<T> &P)
{ for (auto it : P) { s << "<" << it << "> "; } return s; }
template<class T> ostream& operator << (ostream &s, const multiset<T> &P)
{ for (auto it : P) { s << "<" << it << "> "; } return s; }
template<class T> ostream& operator << (ostream &s, const unordered_set<T> &P)
{ for (auto it : P) { s << "<" << it << "> "; } return s; }
template<class S, class T> ostream& operator << (ostream &s, const map<S, T> &P)
{ for (auto it : P) { s << "<" << it.first << "->" << it.second << "> "; } return s; }
template<class S, class T> ostream& operator << (ostream &s, const unordered_map<S, T> &P)
{ for (auto it : P) { s << "<" << it.first << "->" << it.second << "> "; } return s; }
void yes(bool a) { cout << (a ? "yes" : "no") << endl; }
void YES(bool a) { cout << (a ? "YES" : "NO") << endl; }
void Yes(bool a) { cout << (a ? "Yes" : "No") << endl; }
const vector<int> DX = {1, 0, -1, 0, 1, -1, 1, -1};
const vector<int> DY = {0, 1, 0, -1, 1, -1, -1, 1};



// edge class (for max-flow)
template<class FLOW> struct FlowEdge {
    // core members
    int rev, from, to;
    FLOW cap, icap, flow;
    
    // constructor
    constexpr FlowEdge() noexcept = default;
    constexpr FlowEdge(int rev, int from, int to, FLOW cap, FLOW rcap = 0) 
        : rev(rev), from(from), to(to), cap(cap), icap(cap), flow(rcap) {
    }
    void reset() { 
        flow -= icap - cap;
        cap = icap;
    }
    
    // debug
    friend ostream& operator << (ostream& s, const FlowEdge& e) {
        return s << e.from << " -> " << e.to << " (" << e.cap << ", " << e.flow << ")";
    }
};

// graph class (for max-flow)
template<class FLOW> struct FlowGraph {
    // core members
    vector<vector<FlowEdge<FLOW>>> list;
    vector<pair<int,int>> pos;  // pos[i] := {vertex, order of list[vertex]} of i-th edge
    
    // constructor
    FlowGraph(int n = 0) : list(n) { }
    void init(int n = 0) {
        list.clear(), list.resize(n);
        pos.clear();
    }
    void clear() {
        list.clear(), pos.clear();
    }
    
    // getter
    vector<FlowEdge<FLOW>> &operator [] (int i) {
        assert(0 <= i && i < (int)list.size());
        return list[i];
    }
    const vector<FlowEdge<FLOW>> &operator [] (int i) const {
        assert(0 <= i && i < (int)list.size());
        return list[i];
    }
    size_t size() const noexcept {
        return list.size();
    }
    FlowEdge<FLOW> &get_rev_edge(const FlowEdge<FLOW> &e) {
        return list[e.to][e.rev];
    }
    const FlowEdge<FLOW> &get_rev_edge(const FlowEdge<FLOW> &e) const {
        return list[e.to][e.rev];
    }
    FlowEdge<FLOW> &get_edge(int i) {
        return list[pos[i].first][pos[i].second];
    }
    const FlowEdge<FLOW> &get_edge(int i) const {
        return list[pos[i].first][pos[i].second];
    }
    vector<FlowEdge<FLOW>> get_edges() const {
        vector<FlowEdge<FLOW>> edges;
        for (int i = 0; i < (int)pos.size(); ++i) {
            edges.push_back(get_edge(i));
        }
        return edges;
    }
    
    // change edges
    void reset() const {
        for (int i = 0; i < (int)list.size(); ++i) {
            for (FlowEdge<FLOW> &e : list[i]) e.reset();
        }
    }
    void change_edge(FlowEdge<FLOW> &e, FLOW new_cap, FLOW new_rcap) {
        assert(new_cap >= 0 && new_rcap >= 0);
        FlowEdge<FLOW> &re = get_rev_edge(e);
        e.cap = new_cap, e.icap = new_cap + new_rcap, e.flow = new_rcap;
        re.cap = new_rcap, re.icap = new_cap + new_rcap, re.flow = new_cap;
    }
    
    // add_edge
    void add_edge(int from, int to, FLOW cap, FLOW rcap = 0) {
        assert(0 <= from && from < (int)list.size() && 0 <= to && to < (int)list.size());
        assert(cap >= 0);
        int from_id = int(list[from].size()), to_id = int(list[to].size());
        if (from == to) to_id++;
        pos.emplace_back(from, from_id);
        list[from].push_back(FlowEdge<FLOW>(to_id, from, to, cap, rcap));
        list[to].push_back(FlowEdge<FLOW>(from_id, to, from, rcap, cap));
    }
    void add_bidirected_edge(int from, int to, FLOW cap) {
        assert(0 <= from && from < (int)list.size() && 0 <= to && to < (int)list.size());
        assert(cap >= 0);
        add_edge(from, to, cap, cap);
    }

    // augment
    FLOW augment(int s, int t, FLOW up_flow = numeric_limits<FLOW>::max()) {
        vector<bool> seen(size(), false);
        auto dfs = [&](auto &&dfs, int v, FLOW up_flow) -> FLOW {
            if (v == t) return up_flow;
            seen[v] = true;
            for (int i = 0; i < (int)list[v].size(); i++) {
                FlowEdge<FLOW> &e = list[v][i], &re = get_rev_edge(e);
                if (seen[e.to] || e.cap <= 0) continue;
                FLOW flow = dfs(dfs, e.to, min(up_flow, e.cap));
                if (flow > 0) {
                    e.cap -= flow, e.flow += flow;
                    re.cap += flow, re.flow -= flow;
                    return flow;
                }
            }  
            return FLOW(0); 
        };
        return dfs(dfs, s, up_flow);
    };

    // find reachable nodes from node s (1: s-domain, -1: t-domain, 0: no reach)
    vector<int> find_cut(int s, int t) const {
        vector<int> res(size(), 0);
        auto dfs_s = [&](auto &&dfs_s, int v) -> void {
            res[v] = 1;
            for (const auto &e : list[v]) {
                if (res[e.to] || e.cap <= 0) continue;
                dfs_s(dfs_s, e.to);
            }
        };
        auto dfs_t = [&](auto &&dfs_t, int v) -> void {
            res[v] = -1;
            for (const auto &e : list[v]) {
                auto re = get_rev_edge(e);
                if (res[e.to] || re.cap <= 0) continue;
                dfs_t(dfs_t, e.to);
            }
        };
        dfs_s(dfs_s, s), dfs_t(dfs_t, t);
        return res;
    }

    // check if the s-t flow is feasible
    bool is_feasible(int s, int t) const {
        vector<FLOW> b(list.size(), FLOW(0));
        for (int v = 0; v < (int)list.size(); v++) {
            for (const auto &e : list[v]) {
                b[v] += (e.flow - get_rev_edge(e).flow) / 2;
            }
        }
        if (b[s] + b[t] != 0) return false;
        for (int v = 0; v < (int)list.size(); v++) {
            if (v != s && v != t && b[v] != FLOW(0)) return false;
        }
        return true;
    }
    bool is_feasible(int s, int t, FLOW flow) const {
        vector<FLOW> b(list.size(), FLOW(0));
        for (int v = 0; v < (int)list.size(); v++) {
            for (const auto &e : list[v]) {
                b[v] += (e.flow - get_rev_edge(e).flow) / 2;
            }
        }
        if (b[s] != flow) return false;
        if (b[t] != -flow) return false;
        for (int v = 0; v < (int)list.size(); v++) {
            if (v != s && v != t && b[v] != FLOW(0)) return false;
        }
        return true;
    }

    // decompose flow into s-t simple paths and cycles
    using Path = vector<FlowEdge<FLOW>>;
    pair<vector<Path>, vector<Path>> decompose(int s, int t) const {
        struct Arc {
            int to;
            FLOW rem;
            int eidx;
        };
        assert(is_feasible(s, t));
        vector<vector<Arc>> fg(list.size());
        for (int v = 0; v < (int)list.size(); v++) {
            for (int j = 0; j < (int)list[v].size(); j++) {
                FLOW f = list[v][j].icap - list[v][j].cap;
                if (f > 0) fg[v].push_back({list[v][j].to, f, j});
            }
        }
        vector<int> ptr(list.size(), 0), onpath(list.size(), -1);
        vector<pair<int, int>> route;
        vector<int> used;
        vector<Path> paths, cycles;

        auto next_arc = [&](int v) -> int {
            while (ptr[v] < (int)fg[v].size() && fg[v][ptr[v]].rem <= 0) ptr[v]++;
            return (ptr[v] < (int)fg[v].size() ? ptr[v] : -1);
        };
        auto extract = [&](int begin, bool is_cycle) {
            FLOW mi = numeric_limits<FLOW>::max();
            for (int k = begin; k < (int)route.size(); k++) {
                auto [v, i] = route[k];
                mi = min(mi, fg[v][i].rem);
            }
            vector<FlowEdge<FLOW>> seq;
            for (int k = begin; k < (int)route.size(); k++) {
                auto [v, i] = route[k];
                fg[v][i].rem -= mi;
                FlowEdge<FLOW> e = list[v][fg[v][i].eidx];
                e.flow = mi;
                seq.push_back(e);
            }
            if (is_cycle) cycles.push_back(std::move(seq));
            else paths.push_back(std::move(seq));
        };
        auto walk = [&](int start, bool stop_at_t) {
            route.clear();
            int v = start;
            onpath[v] = 0;
            used.push_back(v);
            while (true) {
                int i = next_arc(v), u = fg[v][i].to;
                route.push_back({v, i});
                if (stop_at_t && u == t) {
                    extract(0, false);
                    break;
                }
                if (onpath[u] != -1) {
                    extract(onpath[u], true);
                    break;
                }
                onpath[u] = (int)route.size();
                used.push_back(u);
                v = u;
            }
            for (int w : used) onpath[w] = -1;
            used.clear();
        };

        // extract all s-t paths
        while (next_arc(s) != -1) walk(s, true);

        // decompose remained circulation into cycles
        for (int v = 0; v < (int)list.size(); v++) while (next_arc(v) != -1) walk(v, false);

        return {paths, cycles};
    }

    // debug
    friend ostream& operator << (ostream& s, const FlowGraph &G) {
        const auto &edges = G.get_edges();
        for (const auto &e : edges) s << e << endl;
        return s;
    }
};

// Dinic
template<class FLOW> FLOW Dinic(FlowGraph<FLOW> &G, int s, int t, FLOW limit_flow) {
    assert(0 <= s && s < (int)G.size() && 0 <= t && t < (int)G.size() && s != t);
    FLOW current_flow = 0;
    vector<int> level((int)G.size(), -1), iter((int)G.size(), 0);
    
    // Dinic BFS
    auto bfs = [&]() -> void {
        level.assign((int)G.size(), -1);
        level[s] = 0;
        queue<int> que;
        que.push(s);
        while (!que.empty()) {
            int v = que.front();
            que.pop();
            for (const FlowEdge<FLOW> &e : G[v]) {
                if (level[e.to] < 0 && e.cap > 0) {
                    level[e.to] = level[v] + 1;
                    if (e.to == t) return;
                    que.push(e.to);
                }
            }
        }
    };
    
    // Dinic DFS
    auto dfs = [&](auto self, int v, FLOW up_flow) {
        if (v == t) return up_flow;
        FLOW res_flow = 0;
        for (int &i = iter[v]; i < (int)G[v].size(); ++i) {
            FlowEdge<FLOW> &e = G[v][i], &re = G.get_rev_edge(e);
            if (level[v] >= level[e.to] || e.cap <= 0) continue;
            FLOW flow = self(self, e.to, min(up_flow - res_flow, e.cap));
            if (flow <= 0) continue;
            res_flow += flow;
            e.cap -= flow, e.flow += flow;
            re.cap += flow, re.flow -= flow;
            if (res_flow == up_flow) break;
        }
        return res_flow;
    };
    
    // flow
    while (current_flow < limit_flow) {
        bfs();
        if (level[t] < 0) break;
        iter.assign((int)iter.size(), 0);
        while (current_flow < limit_flow) {
            FLOW flow = dfs(dfs, s, limit_flow - current_flow);
            if (flow <= 0) break;
            current_flow += flow;
        }
    }
    return current_flow;
};

template<class FLOW> FLOW Dinic(FlowGraph<FLOW> &G, int s, int t) {
    return Dinic(G, s, t, numeric_limits<FLOW>::max());
}


//------------------------------//
// Solver
//------------------------------//

int main() {
    cin.tie(nullptr);
    ios_base::sync_with_stdio(false);

    ll N, M, D, INF = 1LL<<40;
    cin >> N >> M >> D;
    vll U(M), V(M), P(M), Q(M), W(M);
    REP(i, M) cin >> U[i] >> V[i] >> P[i] >> Q[i] >> W[i], U[i]--, V[i]--, P[i] -= D;

    vll ts{-INF, INF};
    REP(i, M) ts.EB(P[i]), ts.EB(Q[i]);
    sort(ALL(ts));
    ts.erase(unique(ALL(ts)), ts.end());
    ll T = ts.size();

    FlowGraph<ll> G(N * T);
    ll s = 0, t = N * T - 1;
    REP(i, N) REP(j, T-1) {
        ll v = i * T + j;
        ll v2 = i * T + j+1;
        G.add_edge(v, v2, INF);
        // ll j2 = lower_bound(ALL(ts), ts[j] + D) - ts.begin();
        // if (j2 == j) j2++;
        // if (j2 < T) {
        //     ll v2 = i * T + j2;
        //     G.add_edge(v, v2, INF);
        // }

        //cout << j << " -> " << j2 << endl;
    }
    // REP(i, N) REP(j, T) REP2(k, j+1, T) {
    //     ll v = i * T + j;
    //     if (ts[k] - ts[j] >= D) {
    //         ll v2 = i * T + k;
    //         G.add_edge(v, v2, INF);
    //     }
    // }
    REP(i, M) {
        ll p = lower_bound(ALL(ts), P[i]) - ts.begin();
        ll q = lower_bound(ALL(ts), Q[i]) - ts.begin();
        ll u = U[i] * T + p, v = V[i] * T + q;
        G.add_edge(u, v, W[i]);
    }
    //COUT(G); COUT(s); COUT(t);
    auto res = Dinic(G, s, t);
    //COUT(G); COUT(D); COUT(ts);
    cout << res << endl;
}