Lecture 11 - Supply Uncertainty, Quantity Discounts and Production Lot Sizing

IIT Roorkee July 201
61 followers
1 Video View·Jan 10, 2025
Welcome friends.
In our earlier class, we were talking of how to handle uncertainty in our inventory management.
And in uncertainty related to inventory management, we discussed that largely uncertainties are
in the area of demand side.
That when we are having some kind of uncertainty in consumption rate that is to be handled
and for handling that uncertainty, we have a very important tool which is known as safety
stock.
We discussed that how are we going to calculate the safety stocks.
We discussed some of the measures in the field of uncertainty like product fill rate, order
fill rate and more importantly cycle service labels.
We discussed that how cycle service labels can be done with the help of normal distribution
function in the excel.
Now, you can calculate cycle service label, you can also calculate what should be the
reorder point on the basis of these CSL calculations.
All those things we have done in our last session.
Going further, it is also possible that there may be uncertainties related to supply side
also.
There is a demand side, there is a supply side.
Supply side uncertainties more commonly are possible in two ways.
In all our previous classes, we have assumed that whatever we are ordering, the supplier
is giving the same quantity.
You are ordering 500 units, supplier is giving you 500 units, but you can always assume that
supplier can give short supplies also.
If you are ordering 500 units, supplier may give you just 400 also because supplier has
also done commitments to many customers and only some limited quantity is given to you.
That is quite possible.
Second is supplier is saying you that I am going to deliver you in 10 days.
You give me the order, I will give you in 10 days, but supplier may take 12 days to
deliver.
So, there are uncertainties related to supply side also.
Then in many kinds of inventory issues, when you are procuring the items, there are possibilities
that your supplier may offer you some deal also.
Okay, you are giving me the order of let us say 500 units, I will charge you 5 rupees
per product.
But if you give order of let us say minimum 1000 quantity, I will charge 4 rupee 50 paisa.
And if you give minimum order of 2500 units, I will charge only 4 rupee per unit.
So, these type of situations are also very common in our day to day life.
Even in your personal life also, when you go to vegetable markets, you see these kind
of quantity discount situations, where a vegetable seller, a food seller may charge different
rates if you are going for a bulk buy.
So, if you are getting a situation of quantity discounts, how are we going to change our
decision making in light of that.
And then there are some situations, where you are producing on one side, you are receiving
the order on one side and you are simultaneously consuming them also.
Like in our EOQ model, if you remember, all the replenishments were happening instantly.
A single state vertical line was representing the replenishment line.
But it is possible that replenishment may happen over a period of time and during the
period of replenishment, consumption is also happening simultaneously.
So, these types of models are known as production lot sizing models or economic production models.
So, how are we going to decide the economic order quantity, which we say in this particular
case as economic production quantity EPQ, that is also we are going to discuss in this
particular class.
So, as I said, the first thing which we are discussing that is the supply side uncertainty.
So, if you see this model, here we have this ROP, this is the ROP.
Now, here we are having this fixed lead time, but it is also possible that this lead time
is also variable.
There may be a mean lead time and then there is a standard deviation of lead time.
So, you may have a very complex situation.
You already have a mean demand that we discussed in our last class and standard deviation of
demand and you also have another variable, the mean lead time and standard deviation
of lead time.
So, if you want to calculate the overall demand during the lead time, then you simply
multiply this with this, mean lead time with mean demand.
But if I want to calculate the sigma of demand during lead time, sigma of demand during lead
time, this is calculated with this formula, where if you see this formula, we have two
terms.
One is taking care the variations of the demand and another is taking care the variations
of the lead time, L into sigma D square plus D square SL square, where SL we are representing
as this sigma L t is equals to SL.
So, this is how we are going to calculate the standard deviation of demand during lead
time.
D D L t stands for demand during lead time.
So, these are the calculations and then once you have D L and sigma D D L t, you may recall
what we discussed in our previous class, that normal distribution curve will come into the
picture.
This is D L and then you have one standard deviation, two standard deviation and three
standard deviation.
These are sigma D D L t, this is plus 2 sigma D D L t and this is plus 3 sigma D D L t.
And so on, on the left side of this D L also.
So, here if I have a simple illustration to explain, that how these calculations will
take place.
We have weekly demand of 2500 units, standard deviation of this demand is 500, lead time
is 7 days, L is lead time.
And standard deviation of lead time is also 7 days, the mean lead time is 7 days, this
is in fact actually mean lead time or average lead time and this is the standard deviation
of lead time.
So, our average demand is 2500 that is per day demand into the 7 days, 17500 units.
And then we calculate sigma D D L t using this formula and that sigma D D L t is also
coming to be 17550.
And then we are considering a C S L of 90 percent for this particular situation.
So, if you remember we can use normal inverse function in the excel and that normal inverse
function of this multiplied by sigma D D L t will give you the safety stock 22491.
Now, I have a question for all of you that what should be the use of this safety stock.
You can see that in the bottom line we have mentioned this point that daily demand is
around 2500 units.
If I divide 22491 it is approximately 9 days.
So, meaning is I am keeping a safety stock for handling the situation up to 9 days, up
to the 9 days I can handle the situation and it is possible that there may be some delay
in the supply side also and there may be some increase in demand also.
So, the cumulative effect if I say is possible to handle for let us say 9 days if there is
only variations of the lead time or if there is a increase in daily demand because daily
demand can also have a variation of standard deviation 500 units.
So, in that cumulative manner if 22491 that safety stock is kept I will be able to achieve
on an average a CSL of 90 percent that means 90 percent orders will be fulfilled from the
available inventory in my stocks.
10 percent customers may not be able to get orders from the available stock that is the
impact of supply side and demand side uncertainties.
Then another situation we want to discuss as I said in the beginning that is the quantity
discount.
In the quantity discount situations you are given different price ranges P1, P2, P3, P4
depending upon your order quantity prices will decrease, prices will decrease as your
order quantity is increasing.
So, your order quantity Q is inversely proportional to the price that is the idea of this quantity
discount models.
You charge more price order quantity will be less, you charge less price order quantity
may increase.
Now, here the question is sometime when we want to take the benefit of these quantity
discounts we may not be ordering the EOQ quantities because we want EOQ quantity which
will minimize my total inventory cost.
This price benefit is not coming on the inventory cost, price benefit is coming on the material
cost and if you remember our first class of this inventory discussions our EOQ calculation
is basically based on holding cost and ordering cost.
So, in this case if some discounts are available for larger purchases how to incorporate this
benefit.
Ultimately please remember one more thing though we are discussing many analytical related
decision making and when we started this inventory management discussion we said that we want
to minimize the total inventory cost, but that is not the correct picture.
We want to minimize the overall cost of our operational activities.
Inventory is just one subset of that, but if I am getting a bigger benefit because of
larger purchase quantity then I may sacrifice the benefit of inventory cost.
So, here I have a more holistic model where I will include the material cost as well as
the inventory cost to see that whether the benefits are really significant.
If the benefit because of quantity discounts or bigger purchases can give me some benefit
at the cost of increased inventory cost I will still love to go for those kind of benefits.
But if by increasing the lot size, if by increasing the purchase quantity the benefits are lost
if there is no benefit in fact my total cost is increasing I will not like to go for quantity
discounted purchases.
So, in this way I am going to discuss that how are we going to handle these quantity
discount cases.
In the quantity discount case we have a very systematic process of handling the inventory
calculations.
Here you are having let us say 2, 3 prices we have a requirement of a product which is
coming at the rate of let us say 2000 units per year that is the r.
I am using the same symbols which I started in every session.
So, I am not repeating what is the meaning of r it is 2000 units per year that means
our annual demand.
Now, the supplier is ready to give me 3 different prices one price is rupees 2.
So, you can note down them as rupees 2 for less than 1000, rupees 1.90 for 1000 to 1999
and rupees 1.86 for more than 2000 more than or equal to 2000.
So, now I am getting a quantity related prices that if I am ordering more quantity the supplier
is ready to give me a discounted price.
Now, first and the cost of carrying the inventory is given as 16 percent of average inventory
value this is the data.
Here I have assumed 20 rupees as ordering cost.
Now, as a first step of this particular process I have to calculate different EOQs this is
step 1.
So, I am calculating in this case because 3 prices are given P1, P2, P3.
So, I can say that EOQ1 is basically corresponding to P1, EOQ2 is corresponding to P2 and EOQ3
is corresponding to P3.
With these 3 prices I am calculating EOQs using the same formula under root 2 r Cp upon
Ch.
So, in numerator there will be no change r is constant and all requirement the ordering
cost is also constant 20 rupees per order Ch will vary because 1 is 2 rupees another
Ch is will be calculated for 1 rupee 90 paisa and third Ch will be calculated at 1 rupee
86 paisa.
price.
So, 0.16 into 2, 0.16 into 1.90, 0.16 into 1.86 these will be the values in my denominators.
So, accordingly the calculation tells me 500, 513 and 519 these are the 3 EOQs, but now
you can see if I am charging 1 rupee 90 paisa the minimum order quantity expected is 1000,
but here my EOQ is coming 513 which is less than 1000.
So, the supplier will say hold you will not be getting the benefit of 1 rupee 90 paisa
if you are giving me order of 513 I will charge you rupees 2.
So, he will say I have nothing to do with your EOQ calculations I will charge what is
as per my policy.
So, according to supplier this is invalid EOQ because this is not falling in the prescribed
range for which price rupee 1 price 1 rupee 90 paisa is applicable and similarly our third
EOQ is also invalid, this is also invalid EOQ.
So, both these are invalid EOQs 513 and 519 and only one valid EOQ I could find that is
500 units.
Now, based on these things I will calculate my total inventory cost that is at 3 particular
levels one is at 500 another at 1000 where I am getting a price break and third is at
2000.
The rule we see is that we have to calculate the total cost at all valid EOQs and at all
price breaks.
So, in this case so we have 3 EOQs 500, 513 and 519 these two are invalid only 500 is
valid, valid invalid because of price ranges on which that particular price is applicable.
So, identify only valid EOQs that is 500 and then we have to calculate total cost at
all valid EOQ that is 500 only one valid EOQ and we have two price breaks in this particular
case one price break is at 1000 and another price break is at 2000.
So, in this way we have to calculate the total cost that is 500 EC at 1000 and total
cost at 2000 and that calculations are already available to us on this particular slide at
500 level, at 1000 level and at 2000 level and here we see that these are coming 4160,
3992 and 4038.
I can request all of you to try these calculations on your own you can use excel if you avoid
calculators.
So, you can use excel and see that put the formula and get the values of these total
cost.
Out of these three total cost the minimum cost is 3992 corresponding to quantity level
of 1000.
So, it becomes our order quantity this is our order quantity which is actually giving
the benefit of lowest total cost here we are enjoying some benefits not complete some benefits
of EOQ and the benefit of quantity discount.
So, we have compromised and when I say some benefit of quantity discount I will request
all of you that please explore why it is some benefit you may recall our discussions of
a class where we discussed that at EOQ level the holding cost and ordering cost are equal,
but at this level TCQ equals to 3992 if you calculate the holding cost and ordering cost
they are not equal when we are taking the effect of quantity discount this becomes the
lowest cost and it is the order quantity for this particular case.
Moving further we have another important case that is production lot sizing as I said in
the beginning there may be situations where we are actually in our traditional models
of EOQ these vertical lines are the replenishment lines and these slanted lines are the consumption
lines, but in some cases it is possible that our replenishment lines are like this and
then the consumption lines like this and then the consumption line.
And when we have these kind of angular replenishment lines it means the replenishment and consumption
happening simultaneously the name indicates that only in the manufacturing facility when
we have a production facility we have to discuss about production lot sizing, but any place
where we have replenishment and consumption simultaneously in that case this production
lot sizing is applicable.
Now, the rate of production is P and rate of consumption is R. So, the inventory that
means you are producing let us say 10 units per day and you are consuming at the rate
of 3 units per day.
So, I can say that inventory is getting developed at the rate of P minus R 7 units per day that
is the rate of inventory accumulation.
So, in some time I am running this facility for T days and then I am using the inventory
whatever we have built in these T days over D days and T is smaller than D. T is smaller
than D generally because you are running your facility for 3 or 4 days and then you
are consuming these over a 10 days period or 20 days period.
And one more thing P is the production rate is generally not only greater, but much much
greater than R because if production rate is equal or less than R no inventory accumulation
will take place.
So, these are the two important conditions for developing the production lot sizing or
EPQ model.
Now, here since entire discussion is similar to EOQ model there is only one important change
which you can directly see from this formula that denominator term here we have a multiplying
factor of 1 minus R upon P that is the only change we are having that what is the what
is the rate of accumulation of inventory that is deciding the P level.
You see my production quantity QP is the in one lot I am going to produce QP items, but
my average inventory is not going to be QP level.
If I am producing QP, so QP will be the item I am consuming QP will be the item I am consuming.
So, average inventory in this case is QP by 2, but if QP is not produced over a day QP
is produced over some number of days, then what is going to happen?
I am going to produce same QP, but these QPs are consumed like this.
So, you are producing QP, but you are not touching these peaks.
So, the height of this peak will be this height of the peak will be QP into P minus R upon
2P that is the height of the peak.
And therefore, the average inventory in the case of PQ is QP into P minus R upon 2P that
is the height.
And therefore, this is we already know is the annual ordering cost this will be our
because of this discussion it will be our annual holding cost and that is our QP cost.
So, with this different type of models which are possible in our inventory discussions
and we discussed three important models in this particular class where we discussed that
how you are going to handle the supply side uncertainty.
And then with the help of example, we saw that a complex case where supply side as well
as demand side uncertainties were there.
We also discussed quantity discount case which are very very common in bigger purchases and
then we also discussed that where your replenishment is not instant replenishment is happening
over a period of time and during the period of replenishment consumption is also happening
simultaneously what should be the order quantity and what will be the maximum order level in
these situations.
So, with this we come to end of this session.
Thank you very much.
