this move and to ensure that safety is not compromised. There are already many stories of accidents, injuries and

identified by the risk assessment to ensure safety are implemented.

This differentiation would be okay

replacement for R-134a, widely used in

death related to the inappropriate use of

provided it could be ensured that every

mobile air-conditioners and large water

flammable refrigerants in the commercial

system was installed in accordance with

chillers, is converted almost 100 % to

sector although it is perhaps fair to

the safety standard. Most accidents and

trifluoroacetic acid, but its isomer R-1234ze

say that most of these, if not all, have

injuries occur when the standards are

is more likely to form calcium fluoride,

occurred through ignorance of suitable

not followed. In that case it doesn’ t help

a common and non-toxic mineral which

procedures or malpractice that would

to make the standard more complicated;

occurs naturally in fluorspar crystals. There

not be sanctioned by a competent safety

that’ s just an added burden making

is therefore a case for switching R-1234yf

standard in any case, no matter which

compliance less likely. If flammable

uses to one of the R-1234ze isomers,

refrigerant was being used.

refrigerants are to be used safely there

although they would require larger

Safety standards such as ISO5149,

is an urgent need to improve adherence

compressors for a given cooling duty.

ASHRAE Std15 and EN378 have

to safety standards. Like the electricity

R-1234ze( E) would be a suitable alternative

to R-1234yf, with a similar normal boiling

point, but would require a 30 % larger

compressor. R-1234ze( Z) has a much

higher normal boiling point and also much

traditionally used two methods to promote safe system design:

• The first method, prompted by product standards for air-conditioners, retail equipment or other items, is

generation-cleaning process described above, this has two facets and both are necessary for a successful outcome. The text of the standards needs to be simplified so that it is easy to follow and

Image by © Cold Link Africa

higher critical temperature so it is not

currently used as a refrigerant, but it might

find a place as an alternative working

fluid to R-245fa in organic Rankine cycle

systems, or in high temperature heat

pumps. The four substances are compared

in Table 3, which gives the normal boiling

point and the critical temperature.

Table 3: Key properties of Various

Fluorinated Refrigerants

R-134a-26.1 101.1 40.59

R-1234yf-29.5 94.7 33.82

R-1234ze( E)-19.0 109.4 36.35

R-1234ze( Z) + 9.8 153.7 35.31

If concessions from the PFAS regulations

are not granted for R-32 or the isomers of

R-1234ze then it seems likely that the use

of fluorinated refrigerants will continue

to be increasingly restricted over the

next few years. There will then be few

alternatives in the sectors identified earlier

as the most difficult to adapt.

If hydrocarbons are to be used to fill the

gaps when R-410A, R-32 and R-1234yf are

tightly constrained then there is an urgent

need for all relevant sectors of the industry

to learn how to use them appropriately

and safely. This is a multi-faceted problem

covering component design, system

construction, operation and maintenance.

It also requires a substantial overhaul of

our existing safety standards to reflect

the real hazards that will be created by

to place a strict limit on the amount of refrigerant that can be contained within a unit in a certain location. Typically, this is governed by the volume of the space in which the unit is located and the characteristics of the refrigerant, either in terms of flammability or toxicity.

• The second method, more common for large industrial systems, is to recognise that limiting the amount of refrigerant is not the prime method of ensuring safety and other design features such as gas detection, emergency ventilation, operational procedures and staff training need to be implemented to keep staff safe in the event of a leak.

It is fair to say that each of these methods has its merits and over many years they have a generally good safety track record. However, over the years the first method has become more and more complicated and is increasingly difficult to justify. Recent discussions in the standardswriting community have called these two approaches the‘ intrinsic design method’ and the‘ extrinsic design method’. In the former case, provided the installer purchases a system from a product manufacturer and follows the installation instructions the system will be safe, whereas in the latter case someone must take on the role of system designer for the installation, conducting risk assessments and ensuring that all necessary steps

adoption of the standard needs to be rigorously enforced. There is no point relaxing the standard if nobody uses it, but there’ s also no advantage in enforcing a complex requirement that is impossible to implement in practice.

6. TWENTY YEARS FROM NOW

This overview has taken a broad sweep from the mid-1990s with a 50-year timescale. That would get us to 2045 but it is important at the outset of the future part of the review to emphasise that this is not a final destination for the journey to a better world. Things will continue to change, and new opportunities will continue to appear. The following comments offer an insight into some of the possible futures that may arise.

If the global efforts of the Kyoto Protocol, the Kigali Amendment and the Paris Agreement are as effective in addressing climate change as the Montreal Protocol has been then we can expect to see several themes by 2045. Electricity will largely be sourced from renewables, principally wind and solar power, with the challenge of resilience and intermittency addressed by a range of short-, medium- and long-term energy storage solutions, including pumped hydro storage, central battery banks and the adaption of distributed battery resources, for example in homes or electric vehicles.

South Africa is well placed, provided suitable investment is available, to play a large part in this revolution. Solar power

Dr Andy Pearson.

could be harnessed to manufacture hydrogen to be converted locally to electricity in fuel cells or transformed to ammonia for long-term storage or global transportation. If large thermal plants are required to provide frequency stability for the grid then either hydrogen or ammonia can be burned in turbines, displacing coal and gas to reduce CO 2 emissions.

Wind power can be tapped during periods of high generation capacity to manufacture ammonia for longer term storage, with each 1.5MW turbine requiring a 50m3 ammonia storage tank to provide two weeks of generator output. This ammonia storage could be held locally, or could be centralised, depending on the resilience of the distribution grid. In this sense, the target of 30 % renewables by 2030 should be seen as the start of the process, not the conclusion.

Having‘ greened the grid’, more and more users of heat and cold will transition onto the electrical network, including transport and a much-extended cold chain. The population of Africa will have increased by almost one billion people by 2045 and most of the additional number will be living in cities not rural communities.

The growth in the cold chain to serve that population will be substantial and to avoid placing excessive strain on the infrastructure, refrigeration systems will