Apple's latest analysis of its products' lifecycle CO2e emissions show that manufacturers can have a biggest impact on ICT's carbon footprint.

'Embodied' carbon is the carbon emission associated with a piece of ICT gear before the user first switches it on. Apple's 2018 Environmental Responsibility Report presents an analysis of it's products' lifecycle in aggregate.

77% of total lifecycle CO2e are generated in the Manufacturing phase, with integrated circuits being the biggest contributor. "The CPU, SOC, DRAM, and memory (NAND) make up the bulk of the integrated circuit emissions."

By contrast, Product Use generates 17% of lifecycle emissions. This suggests that manufacturers have 4.5x leverage in reducing gear lifecycle emissions compared to users.

The balance of the footprint comes from Product Transportation (4%), Corporate Facilities (1%), and End of Life (1%).

How does this compare to the Apple data we analyzed eight years ago?

Apple reports that its average CO2e per product dropped from 137 Kg in 2011 to 90 Kg in 2017. This is a significant drop, but the data below suggests that much of it could be attributed to a shift in product mix toward smartphones. Even back in 2010, iPhones had lower lifecycle CO2e footprints than today's product average. The mix shift to phones has given Apple cover to increase the footprint of iPhones while claiming its lowering average product footprints.

User CO2e ranged from 28% to 59% of lifecycle footprint in 2010 for non-server products, as displayed below. Since the average product footprint has dropped 34%, the significant drop in the use portion suggests that much of the products' overall drop has come from the product use side, rather than the manufacturing side. In fact, Apple states, "Since 2008, we’ve reduced the average energy consumed by Apple products by 68 percent."

2010 Data

Edge Gear

Table 1 makes two important points critical to Green ICT.

First, we see reinforcement for the fact that most edge gear's environmental impact occurs before it even arrives in our facility or home. This is true of for all but one of these devices just on the basis of the CO2e alone ('embodied' CO2e). Add in the other impacts of materials extraction, manufacture, and transportation and the pre-use impact can reach 75% of life-cycle impact for some gear.

Second, compare the impacts between the two desktops (Mac mini and Mac Pro) or between the two laptops (MacBooks). We see the environmental cost of buying the most powerful option available, something we have previously noted for storage devices and networking gear.

More on CO2e from Apple edge gear in our posts on the Mac mini and the iPhone4.

Configs Matter

We discovered evidence of a third Green ICT principle while mining the Apple CO2e data: configuration details and use cases can can cause energy/CO2e to vary widely. Comparing Table 2's A and C configurations illustrates how the attachment of significantly higher energy/CO2e components can undermine a nicely-performing base configuration. Comparing A and B illustrates how deploying the same hardware** for different uses can yield very different results.

For another example, see how the addition of Apple's high-bandwidth networking cards disqualifies the Mac Pro from EPEAT Gold status.

Size Matters

Finally, there is the "how big and powerful a machine do you need?" issue. Compare the opposite ends of Apple's notebook line in Table 3. Most laptop lifecycle CO2e comes from manufacture/disposal, not use. Prematurely replacing an older laptop with a greener one isn't greener. As with much ICT gear, a bigger, more powerful laptop carries a substantial CO2e premium.

Servers

Servers are the opposite of laptops. Look at configuration B in Table 2, the Mac mini server. Note that its usage CO2e is a whopping 77% of lifecycle CO2e. A Microsoft report suggest it could be as high as 91%: "Hardware comes with an 'embodied' carbon footprint from the energy associated with producing, distributing and disposing of equipment. For the scenarios analyzed, this energy outlay adds about 10 percent to the footprint from IT operations."

It may make more sense to replace an inefficient server than an inefficient laptop. Microsoft claims a switch to cloud computing offers an even more efficient alternative than server replacement.

Apple is to be commended for publishing this data. You can review the Environmental Reports for most Apple products online.

* Does not include CO2e associated with recycling, which is typically an additional 1%-2% of the total CO2e. Use CO2e is for a four-year period.

*** A and C are running Snow Leopard; B is running Snow Leopard Server. Mac mini with Snow Leopard Server is an ENERGY STAR qualified small scale server.