How to cook beans

I have a dirty little secret to tell you -- I often use canned beans.  Most chefs and cooking instructors would find this to be an unforgivable sin, considering that fact that my job is to teach people how to cook.  Surely, I am supposed to tell you to stop buying beans in cans and to start cooking your own -- not only is it cheaper, but a lot tastier too.  At least that's the theory.  Unfortunately, I can't bring myself to give you a lecture on the virtues of cooking your own beans even though I hope to solve your bean cooking problems by the end of this post.  Beans are one of the most temperamental ingredients I've ever dealt with.  They often turn out either too chalky, or too mushy, or their skins burst making their insides taste all watery.  The worst part is that all three problems can happen in one batch making them very difficult to solve.

You didn't expect such an encouraging introduction, did you?  Well, here is the good news.  It is possible to cook perfect beans at home, but it takes lots of patience and precision.  First, let's talk about when you should bother to cook beans yourself and when you can use canned beans with very good results.

Salads
When I need 1-2 cups of beans for a salad -- they are great tossed with tomatoes, or blanched asparagus or green beans -- I usually open a can.   I always keep 3-4 cans of cannellini beans (my favorite type) on hand for just this purpose.  The brand of a can makes a big difference.  Whole Foods generic brand "365" is great.  Just avoid any "No Salt Added" cans.  Beans need salt not only for the purpose of seasoning, but also for even cooking.  Rinse the beans thoroughly in a colander, drain, and serve.  Yes, you'll probably spend $1 on a cup of cooked beans instead of 25 cents if you were to buy them dry.  But when you take your time into account, canned beans will beat home-cooked by a huge margin.

Soups and Stews
When cooking a bean soup or stew, I usually cook my own beans because I want their lovely cooking liquid. The liquid surrounding canned beans is starchy and unpalatable.  But when you cook the beans yourself, you end up with a very flavorful broth that can be used in soups, stews, and sauces.

Spreads
When using beans for a pureed spread (hummus, cannellini rosemary spread, or bean walnut pâté), I often cook my own to have the cooking liquid on hand to adjust the consistency.  Of course, you can also use canned beans and simply add a little water or olive oil.  Pureed bean spreads are also a good "recovery" dish for those cases when your beans refuse to behave.  Just make sure they aren't chalking and crunchy, but don't worry if they burst or end up mushy.  Once you puree them, no one will notice the difference.

If you are going through the trouble of cooking beans, cook a big batch and use them in different dishes.  They'll last happily in the fridge for 1 week when stored in their cooking liquid.

Now that you know when to bother cooking your own beans, let's talk about how.

Bean types
You can use this method for all kinds of white beans (cannellini, navy, great northern), pinto beans, black beans, cranberry beans, black-eyed peas, garbanzo beans (chick peas), and kidney beans.  This method does not apply to lentils.  Although they are also of the legume family, they don't need soaking, and taste best when cooked al dente, while the beans taste best when cooked till completely soft.

Buying beans
Try to buy your beans in stores with a good turn around.  Old beans will not cook evenly and sometimes will refuse to get tender.  Whole Foods usually has a better turn around for beans and grains than most supermarkets.

Soak in salted water (8-24 hours)

To soak 1 Lb of beans, use 2 quarts of filtered or bottled water and 3 Tbsp Diamond Crystal Kosher salt (or 1 Tbsp and 2 tsp of table salt). If you can't cook the beans after soaking for 24 hours, drain, put in a zip lock bag and refrigerate for up to 4 days. If you have a scale, here is a much easier way to measure salt that doesn’t depend on the salt type. Basically, you need a 1.5 – 3% brine by weight. I prefer to use 3%. So for 1 Lb of beans, you need to weigh 1500 grams of filtered or bottled water and 45 grams of salt (any salt is fine, just avoid Table salt since it contains iodine).

Picking through the beans

After the beans have been soaked, drain and rinse them. Drop them onto a hard surface, like a ceramic plate, a handful at a time. Any that make a “ding” sound should be discarded when cooking thin skinned beans like cannellini.  If they are hard as rocks after soaking, they won’t soften after cooking either. Any that make a hollow sound are fine.

Cooking
1 Lb dried beans, soaked as described above, drained, rinsed, picked through
3 quarts filtered or bottled water
1 Tbsp Diamond Crystal Kosher salt (or 1.5 tsp table salt) – skip if using a 3% brine

Optional aromatics:

1 onion, cut in half
1 carrot and 1 celery stick but into large chunks
Bay leaf
A few sprigs or rosemary, sage or thyme

Put beans, water, salt, and any optional aromatics in a large, heavy pot. Bring the beans to a simmer over high heat. As soon as the bubbles break the surface of the water, turn down the heat to very low (if using an electric stove, temporarily move the pan off the burner to let it cool off). Skim off the foam that rises to the top. From here on, you want to keep the liquid around 200 - 207F (that's just under a simmer). No, it's not too anal to use a thermometer every so often to make sure you are in that zone.

I wish I could give you some nice chart with cooking times for different bean types. Unfortunately, I'd be just setting you up for failure. Each batch of beans has a mind of its own, so you have to taste, taste, and taste some more. I start tasting 45 minutes after the beans come to a simmer. Usually, they are done 1-2 hours after they come to a simmer. If they are not done after 3 hours, you are dealing with old beans and it’s time to give up on this batch.

Tasting
Don't be alarmed if the bean skins blister and break when you remove them from the cooking liquid to taste. That's not an indication that they are overcooked. Once the beans have cooled, the skins will become much more stable. Test the beans whose skin looks intact. The ones that broke down a bit are definitely done. If the bean you tasted is completely tender, taste at least 5 more beans to make sure they are all cooked. When working with cannellini and a few other white bean varieties, don't get alarmed if about 10-20% of the beans break slightly. That's usually what happens by the time all of them are tender. Just let them sit in the fridge overnight and their texture will become much more pleasant.

Cooling
Never drain and use the beans immediately after cooking. Their skins often blister and burst unless you give them a chance to cool first. Cool and store the beans in their cooking liquid. Once they cool off to room temperature, refrigerate for up to 1 week. They always taste better the day after they were cooked.

Pressure cooker method
Put 1 Lb soaked, drained, rinsed and picked through beans in a pressure cooker. Add 1500 g filtered or bottled water (that’s about 7 cups). Cover, lock, and bring to pressure on high setting (15 psi). Turn down the heat and cook for 5 minutes for most beans (10 minutes for chickpeas). Take off heat and let sit exactly 10 minutes. Release the steam, unlock, and test for doneness as described above. If not done, continue to cook without pressure uncovered at a bare simmer until done. For more info, consult the Hip Pressure Cooker bean chart.

Lentils braised in red wine

Judy Rodgers describes using a risotto method for lentils in her Zuni Cafe cookbook.  I found the idea intriguing and have been playing with it for over a month now with great success (hey, what recipe from the Zuni cookbook is not a great success?!)  The recipe asks you to add liquid a little at a time, like you would for a risotto.  This stingy with liquid method results in lentils that are nutty and toothsome, coated in a syrupy red wine sauce.  With a bit of practice, I found that I can minimize stirring by adding just the right amount of liquid and even use my oven risotto method eliminating stirring all together.

The recipe starts with cooked mirepoix (mixture of carrots, celery, and onions), but this step is optional.  If I am planning to combine the lentils with caramelized onions later, I skip the mirepoix.  

What I love the most about this method is how forgiving it is.  Cooking lentils in a lot of liquid can yield nice toothsome results as well, but you have to watch your lentils like a hawk.  Overcook them by 2 minutes and they turn to mush.  With the risotto method, the lentils hold their shape and texture a lot longer after they are done.  

Type of lentils

The type of lentils is very important.  You need du Puy (a.k.a. French Green or just French) lentils for this dish.  Black beluga lentils will also work, but avoid brown, yellow, and red lentils since they won't hold their shape as well.

Serves 4 as a side dish

3 Tbsp olive oil

1 yellow onion, diced small

1 medium carrot, diced small

1 celery stick, diced small

1 1/4 c du Puy lentils, rinsed and drained

1 cup medium body, low tannin red wine (I like to use Cote-du-Rhone or Pinot Noir)

1 1/2 cup water, plus more as needed

3 sprigs fresh thyme

1 bay leaf

1-2 Tbsp butter

Salt and pepper to taste

Cooking aromatic vegetables (this step is optional)

Set a medium sauce pan over medium-low heat.

Add the oil, onions, carrots, and celery.  Sprinkle with a generous pinch of salt and cook stirring occasionally until the vegetables are tender, but not brown, 10-15 minutes.  

Cooking lentils -- stove top method

Add the lentils, wine, water, thyme, bay leaf and salt (1/2 tsp Diamond Crystal Kosher or 1/4 tsp table).  Bring to a simmer and cook stirring occasionally at a gentle simmer until most of the liquid is absorbed.  You don't need to stir much when the liquid is plentiful.  As it gets low, start stirring more often.  Taste the lentils.  If they are not tender enough, add more water 1/2 cup at a time and continue to cook stirring and tasting.  The total cooking time will be around 30 minutes.  

Cooking lentils -- oven method

Preheat the oven to 400F.  

Add the lentils, wine, water, thyme, bay leaf and salt (1/2 tsp Diamond Crystal Kosher or 1/4 tsp table).  Bring to a simmer, cover and place in the oven for 30 minutes.  Taste the lentils.  If they are not tender enough, add more water 1/2 cup at a time and continue to cook on the stove top stirring and tasting.  

Seasoning and enriching (both the stove top and oven methods)

When the lentils have the desired texture.  Take them off the heat.  Add more salt as needed and 1-2 Tbsp butter.

More about thermometers -- specifically when and how to use them

Dear readers,

Thank you so much for your thoughtful comments about meat thermometers.  I wanted to reply to all of you, but didn't want to bury the interesting bits in the comments, so I am writing another post.

First of all, I'd like to apologize for my laziness in the last post.  It was really silly of me to start discussing pros and cons of different thermometers without talking about the tasks people do with them.  Considering the fact that my last career was in the field of usability engineering, that's an unforgivable sin.  The only thing I have to say in my defense is that I've been a little distracted the last 2 weeks with a certain fantasy book series.  If I was 14 years old, I wouldn't be embarrassed to tell you the title.  But since those days are long gone, there is really no excuse for my juvenile behavior.  So, I will try to be good now and concentrate.

The task some of you have mentioned when praising probe thermometers was cooking a large prime-rib roast.  I can see how the probe can seem very attractive.  You don't need to constantly open the oven and check your meat.  The thermometer will tell you the exact moment when it's done.  Right?  Not exactly.  Here are some ideas to think about.

Opening the oven is not all that bad
Nothing terrible will happen to your meat if you open the oven and check up on it.  Even if you do it frequently.  You don't want to stick a thermometer in and wait 25 seconds to get a good reading while the oven is opened.  Quickly get the meat out and close the oven door.  Test the meat and then return it to the oven.  If anything, this slight cooling off period will result in a more evenly cooked meat.

You still need to have a rough idea of how long the roast will take
If you don't have any idea when the meat is getting close, how would you know when to put it in the oven to have it ready by the time your guests arrive?  Also, how would you know when to start the end game with your side dishes?  In real life, you are not just sitting there waiting for the beep from your thermometer.  You should probably know roughly when the meat will be ready and use the thermometer often in the last stages of roasting to get perfect doneness.  That's very hard to do for large pieces of meat, which brings me to the last point.

You don't need to cook a huge roast even when feeding lots of people
I never cook a prime rib roast bigger than 3 ribs.  If I am feeding a crowd, I cook several such roasts.  It's much easier to estimate when a little roast will be ready and it's much easier to stick it in a skillet in the end for a good crust.  Here is my general approach to cooking all meat.  Ideally, you should also do a dry-run before the big event to get a rough idea of timing.  If you underestimated how long it will take, your guests might have to wait another 10-15 minutes, not another hour.  There is no reason why regular digital thermometers wouldn't do a good job with a prime rib task even though they might require a bit more work on your part.

But let's talk about tasks that happen much more frequently, at least in my kitchen:

• testing steaks 
• testing duck breasts
• testing pork chops
• testing racks of lamb
• testing chicken breasts and thighs

The common theme here is pieces of meat that are 1 inch in thickness or more.  Let's use a thick steak as an example.  It's very rare that I cook only one steak, and it's very rare that the other ones are exactly the same thickness, and it's very rare that I managed to stick a thermometer smack into the center of the steak on the first try.  What does all this mean?  It means that I need to be able to remove the thermometer easily and stick it into another piece of meat or into another spot on the same piece of meat.  That's where a remote probe thermometer can turn out to be a hindrance.  It gets hot in the oven and becomes very awkward to pull out and stick back in since you need to use oven mitts.  The wire attaching the probe to the thermometer causes some difficulties too since it often makes the thermometer fall down as the probe is yanked out of the meat.  Maybe I am just a very un-coordinated person, but I find regular thermometers to be a lots easier to deal with for this task.

Another task you guys mentioned was monitoring the temperature of water for cooking beans in the oven.  You got me there.  I've never tried it, so I can't say for sure how a probe thermometer would stack up against a regular one.  But I wonder how much good does it do to know the exact temperature of the liquid if you can't easily control it.  Ovens cycle on and off, so they can't give you such perfect control.  What I might do is bring a pot of beans to 185-195F on the stove top while monitoring it with a regular thermometer.  Then set it in 200F oven and hope for the best.

I was getting concerned about my bias against probe thermometers.  I thought maybe I just had a few bad experiences with them, so I decided to check what Cook's Illustrated had to say:

We tested 11 models-several by the same manufacturers-and not one was flawless. The ones that accurately measured temperature sported function buttons that were too slow or too hard to figure out. Others that were user-friendly were also unreliable. 

The best of the bunch (ThermoWorks Original Cooking Thermometer/Timer) was great when it worked but has probes that even its manufacturer admits are sometimes defective. Until a better meat probe comes on the market, we recommend this one with reservations. Check the probe's accuracy by boiling water and taking a reading before trying it with a roast. If the probe doesn't read very close to 212 degrees, ask for a replacement.

It seems to me that the probe thermometers give people a false sense of security.   If you love the results you are getting with them, that's great.  But if not, it might be worth looking at a regular thermometer rather than another probe type.

Meat thermometers: should you spend $10 or $100?

How much should you spend on a meat thermometer?

I bought my first meat thermometer 9 years ago.  It was a fancy one with remote probe and cost about $40 (they sell for $20-30 now).  The idea was that you could stick the probe into the meat, clip the thermometer on the door of the oven, and program it to beep when the meat reached a certain temperature.  At least that's how it worked in principle.  Here is what happened in practice.  The probe didn't stay in the center unless it was a really big hunk of meat.  The wire made the probe awkward to work with and wash.  The probe broke after less than a year of use since it was always jammed by oven door and grill cover

I tried a few more thermometers of this type, but had the same problems, and an added one of the user interface getting more and more complex.  

Finally, I gave up and switched to a basic Taylor digital thermometer that you can't put in the oven.  It cost around $15 (now it sells for $10).  I loved it.  It was easy to stick into the meat, easy to read, and easy to wash.  This thermometers has lasted me close to 5 years now, but you never know when you might drop it into water, so I bought 2 back ups -- Maverick (recommended by Cook's Illustrated) and Tru Temp (available at Target).  

A few weeks ago, I noticed some discrepancy in my thermometers while testing a roast chicken recipe.  Maverick and Taylor where within 1 degree, but Tru Temp (made by Taylor by the way) was 15 degrees off!  15 degrees is a lot.  It was particularly disconcerting since I started using these thermometer for sous-vide preparations and needed accurate readings.  I threw Tru Temp away, but now I was starting to doubt my other thermometers.  After all Tru Temp and Taylor were made by the same company.  

That's when I finally decided to splurge on Thermapen.  $100 did sound a little freaky.  I don't believe in expensive kitchen equipment and rarely spend more than $50 on any kitchen toy.  But with Jason's encouragement, I finally gave in and bought it.  The guy has a thing for lab equipment, so he couldn't help himself.  What can I say -- it's good and it's fast.  I can get a close enough reading in 3 seconds rather than 20 I am used to waiting.  It also allowed me to test my Maverick and Taylor thermometers for accuracy and speed of coming to temperature.  I was testing them all on 130 degree water.  All three thermometers were within 1 degree of each other.  The main difference was in how long it took to get a stable reading.  Thermapen (the Ferrari of all thermometers) was of course the fastest.  Taylor was a lot slower and Maverick was a snail.

• Thermapen -- within 2 degrees after 3 seconds, stabilized after 7.
• Taylor -- within 2 degrees after 17 seconds, stabilized after 20.
• Maverick -- within 2 degrees after 22 seconds, stabilized after 30.

But wait!  Don't spend $100 on your thermometer just yet.  Here is a little practical matter I haven't considered until my last meat class.  A thermometer is only as good as your ability to insert it into the very center of the meat. Just stick it 1/4 of an inch too high or too low and it will throw your reading off.  It's not like the other thermometers don't have the same problem.  They do.  But is it worth paying $100 for perfect accuracy and speed when there is such a huge margin for human error?  

It depends on what you are using your thermometer for.  For testing meat and poultry, I'd stick with the $10 thermometers.   I would just suggest buying 2 different brands and staying away from Tru Temp, so that you can check them against each other once or twice a year.  The most important thing to remember when testing meat is to insert the thermometer sideways.  If you reached the desired temperature, insert the thermometer into 2 more spots trying to aim for the center of the meat to make sure that what you got was indeed the minimum temperature.  The same principle applies no matter how accurate your thermometer is.  

But if you are using your thermometer for sous-vide, thermapen might be worth it.

Here is a continuation of this discussion.

Hazelnut Buckwheat Yogurt cake

What do you do with your 3 year old when it's too cold to play outside?  You bake, of course.  There must be thousands of blog posts devoted to this topic. I will spare you yet another cute story of a 3 year old learning to break eggs and mix batter.  Instead, I'll talk about the recipe.  Here were my requirements when finding the master recipe for our Friday baking sessions with Sammy

1) all ingredients need to be normally found in my pantry and fridge
2) no ingredients at room temperature -- with a toddler and a baby, I never know if and when baking can happen, so I need a recipe that is very forgiving
3) only 2 bowls -- one for wet and one for dry
4) no separating eggs
5) can be baked as either cake or muffins
6) lends itself nicely to variations

The yogurt cake from Chocolate & Zucchini seemed to fit all the requirements, so Sammy and I gave it a shot.    The first time we made it almost as is, replacing about 10% of AP flour with buckwheat and adding apples and blueberries.  It was great and leftovers were excellent for breakfast.

The second time, we were a little more adventurous and replaced 20% of AP flour with buckwheat and 20% with hazelnut meal.  This time we baked it as muffins stuffing some with pears and some with Nutella.  Both were yummy, but I have a soft spot for buckwheat chocolate combination and it looks like Sammy inherited this preference.  The only thing I would do differently next time is add Nutella after the cake or muffins are baked. It doesn't melt particularly well when added as a small lump before baking.  I am guessing, we could put it in a zip lock bag, cut the corner, and pipe like icing on top.  Bet Sammy would enjoy that.

Wet
2 large eggs
227g (1 cup) whole milk plain yogurt (we used Stonyfield)
200g (1 cup) sugar
70g (1/3 cup) vegetable oil
1 tsp vanilla extract
1 Tbsp rum or brandy

Dry

260g (2 cups spooned and leveled) all-purpose flour (or 60g buckwheat, 60g hazelnut, and 140g all-purpose)

1 1/2 tsp baking powder

1/2 tsp baking soda

a good pinch of salt

• Preheat the oven to 350F.  
• Line the bottom of a round 9 or 10 inch cake pan with parchment paper and grease the sides.  Or grease 12 muffin cups (about 1/2 cup each).
• In a medium bowl (I use a 4 cup pyrex measuring cup), mix the wet ingredients.
• Sift dry ingredients into a large bowl.  
• Add the wet ingredients to the dry and fold gently just until no flour streaks remain.  Do not over-mix (a few little lumps are ok).
• Pour the batter into prepared baking pan.  If you want, push some fruit into it (apples, pears, plums, apricots, peaches, blueberries, raspberries, etc).  Optionally, sprinkle with some rolled oats and/or demerara sugar.
• Put in the oven until the toothpick inserted in the center comes out clean.  Here are some rough baking times.  40-45 minutes for a 9 inch pan, 30-35 minutes for a 10 inch pan, 20-25 minutes for muffins.
• Let stand 10 minutes.  Run a knife around the outside to dislodge and transfer to a cooling rack.