With $9$ degrees of freedom? You should expect a number in the range 5 to 20, say. A big values like this means the "null hypothesis" does not fit the data well.
Edited, after checking your calculations and rereading your comments. I get the same chi-squared value. It means, your digit counts do not fit the $\log_{10}(1+1/d)$ formula.
As you have guessed, a large sample size will make the test statistic bigger, if the model is wrong or even just slightly wrong. (You can see this if you artificially multiply all the counts by, say, 1000. Then the test statistic will also be multiplied by 1000, too.)