No need to change anything if you leave out the nuts. Was the recipe from a cookbook? I use 1-1/4 teaspoon of baking soda and 1/4 teaspoon of baking powder in my cakes but it's not the same as yours.

Your recipe depends on both the baking powder and the baking soda to get enough rise. The baking soda works by reacting with the acid in the buttermilk. The baking powder needs to be fresh, and it is the most likely cause of the (occasional) problem. Always keep it tightly covered, and replace it before the expiration date.
The proportion of sugar to flour is high. If I were making this cake, I would cut the sugar. And use butter instead of at least some of the shortening.

The recipe calls for cake flour -- is that what you always use? I've never used cake flour and wonder if its [lower] gluten is the cause of occasional rising problems...

Something I learned with improvisational cake baking under less than ideal circumstances is that so long as you put in the baking powder, no matter the exact quantity, you'll get cake. The amount listed should be fine, plus there's the soda/buttermilk.

Are you using cane sugar? I've recently learned how bad beet sugar is in cookies, and it might be similarly bad in cakes.

I agree that double acting baking powder would help, having the heat component to the rise.

I would add ...are you bakign it long enough? Not by time but are you testing the cake for done, and is your oven properly hot before you put it in the oven and are you baking the cake immediately after mixing...not letting it sit while the oven heats up before you put the pans in the oven.

To reiterate what Linda said, since the cake 'sometimes' works, I strongly suspect that the oven you're using may be the culprit. Make double sure that your oven is properly preheated. I had an oven that required at least 30 minutes to preheat...a big reason why I love my counter-top oven for smaller tasks. I have one that is large enough to fit a 9x13 pan and use it at least 5 x a week.