Minimizing variance - can't seem to find the solution


I have a dataset that consists of Price and Quantity.
The industry has a price that changes every hour or every day, and is calculated off a Supply and Demand formula where both the supply and demand is constantly changing. The Demand part of this is the Industry demand for that hour of that day, but suppliers may be coming and going throughout the day with different quantities and price offers.

As a result there can be large fluctuations in Price between months, both from an overall average price perspective, as well as fluctuations within the day.

I have to pay for my customers' usage at the Industry Price, but their load isn't necessarily that correlated with the Industry Load. There is a big difference if a customer is a 5 or 7 day a week operation, or if it is a 9am-5pm operation vs a 24hr operation etc.

I can only buy hedges at a flat hourly quantity and price for each month. This means I am exposed during the the 'peak' usage periods through the day which is normally in the mornings from around am to 10am and the evenings from around 6pm to 9pm. It also means I am usually long over the weekend and at nights.

I am wanting to offer prices to my customers in 4hr block , split into a weekday/weekend split that changes each month for the next 2 years.
But I am wanting to minimize the variance between what I pay the Industry and what I charge the customer, but am struggling to think of the best approach to create the pricing ratios to apply.

Any help would be most appreciated.

I have provided a sample of a couple of days data for one specific customer which hopefully clarifies some of my words above.