\emph{$K$-best enumeration}, which asks to output $k$-best solutions without duplication, is a helpful tool in data analysis for many fields. In such fields, graphs typically represent data. Thus subgraph enumeration has been paid much attention to such fields. However, $k$-best enumeration tends to be intractable since, in many cases, finding one optimum solution is \NP-hard. To overcome this difficulty, we combine $k$-best enumeration with a concept of enumeration algorithms called \emph{approximation enumeration algorithms}. As a main result, we propose a $4$-approximation algorithm for minimal connected edge dominating sets which outputs $k$ minimal solutions with cardinality at most $4\cdot\OPT$, where $\OPT$ is the cardinality of a minimum solution which is \emph{not} outputted by the algorithm. Our proposed algorithm runs in $\order{nm^2Δ}$ delay, where $n$, $m$, $Δ$ are the number of vertices, the number of edges, and the maximum degree of an input graph.