#pragma GCC optimize("O3")
#pragma GCC optimize ("unroll-loops")
#pragma GCC target("avx,avx2,fma")

#include "bits/stdc++.h"

using namespace std;

#define pb push_back
#define F first
#define S second
#define f(i, a, b)  for(int i = a; i < b; i++)
#define all(a)  a.begin(),a.end()
#define rall(a) a.rbegin(),a.rend()
#define sz(x) (int)(x).size()
#define mp(x, y) make_pair(x,y)
#define popCnt(x) (__builtin_popcountll(x))
#define int ll

using ll = long long;
using ii = pair<int, int>;
using ull = unsigned long long;

const int N = 5e5 + 5, LG = 18, MOD = 1e9 + 7;
const long double PI = acos(-1);
const long double EPS = 1e-7;

#define rep(aa, bb, cc) for(int aa = bb; aa < cc;aa++)
struct Dinic {
	struct Edge {
		int to, rev;
		ll c, oc;
		ll flow() { return max(oc - c, 0LL); } // if you need flows
	};
	vector<int> lvl, ptr, q;
	vector<vector<Edge>> adj;
	Dinic(int n) : lvl(n), ptr(n), q(n), adj(n) {}
	void addEdge(int a, int b, ll c, ll rcap = 0) {
		adj[a].push_back({b, (int)adj[b].size(), c, c});
		adj[b].push_back({a, (int)adj[a].size() - 1, rcap, rcap});
	}
	ll dfs(int v, int t, ll f) {
		if (v == t || !f) return f;
		for (int& i = ptr[v]; i < (int)adj[v].size(); i++) {
			Edge& e = adj[v][i];
			if (lvl[e.to] == lvl[v] + 1)
				if (ll p = dfs(e.to, t, min(f, e.c))) {
					e.c -= p, adj[e.to][e.rev].c += p;
					return p;
				}
		}
		return 0;
	}
	ll calc(int s, int t) {
		ll flow = 0; q[0] = s;
		rep(L,0,31) do { // 'int L=30' maybe faster for random data
			lvl = ptr = vector<int>((int)q.size());
			int qi = 0, qe = lvl[s] = 1;
			while (qi < qe && !lvl[t]) {
				int v = q[qi++];
				for (Edge e : adj[v])
					if (!lvl[e.to] && e.c >> (30 - L))
						q[qe++] = e.to, lvl[e.to] = lvl[v] + 1;
			}
			while (ll p = dfs(s, t, LLONG_MAX)) flow += p;
		} while (lvl[t]);
		return flow;
	}
	bool leftOfMinCut(int a) { return lvl[a] != 0; }
};

int n, m, t;
char grid[22][22];
const int dx[] = {-1, 1, 0, 0};
const int dy[] = {0, 0, -1, 1};

int getId(int i, int j, int t, bool o) {
    int val = n * m * t;
    val += i * m + j;
    return val * 2 + o;
}

void doWork() {
    cin >> n >> m >> t;
    f(i, 0, n)f(j, 0, m) cin >> grid[i][j];
    Dinic flw(n * m * (t + 1) * 2 + 2);
    int src = n * m * (t + 1) * 2;
    int snk = n * m * (t + 1) * 2 + 1;
    f(i, 0, n)f(j, 0, m) if (grid[i][j] != '#') {
                if (grid[i][j] == 'P')flw.addEdge(src, getId(i, j, 0, 0), 1);
                if (grid[i][j] == 'E') {
                    f(x, 0, t + 1)flw.addEdge(getId(i, j, x, 1), snk, 1);
                }
                f(x, 0, t + 1)flw.addEdge(getId(i, j, x, 0), getId(i, j, x, 1), 1);
                f(x, 0, t)flw.addEdge(getId(i, j, x, 1), getId(i, j, x + 1, 0), 1);
                f(k, 0, 4) {
                    int ni = i + dx[k];
                    int nj = j + dy[k];
                    if (ni >= 0 && nj >= 0 && ni < n && nj < m && grid[ni][nj] != '#') {
                        f(x, 0, t) {
                            flw.addEdge(getId(i, j, x, 1), getId(ni, nj, x + 1, 0), 1);
                        }
                    }
                }
            }

    cout << flw.calc(src, snk) << '\n';

}

int32_t main() {
#ifdef ONLINE_JUDGE
    ios_base::sync_with_stdio(0);
    cin.tie(0);
#endif // ONLINE_JUDGE

    int t = 1;
//    cin >> t;
    while (t--) {
        doWork();
    }
}