SELECTING A

PHOSPHORIC ACID
PROCESS
Di, Hemi, & Hemi-Di Processes
for New Plants & Conversions

John Wing, P.E.

American Institute of Chemical Engineers

Clearwater Convention

Clearwater Beach, Florida, USA

June 7, 2008

Content of this paper Copyright by John W ing, P.E.

SELECTING A PHOSPHORIC ACID PROCESS

Di, Hemi, & Hemi-Di Processes for New Plants & Conversions
by John Wing, P.E.
AIChE Clearwater Convention, 2008

SUMMARY
Most of the worlds phosphoric acid is produced by either a dihydrate (Di), hemihydrate
(Hemi) or the hemi-dihydrate (Hemi-Di) process. Which process is best for that new
phosphoric acid plant? Should an existing plant be converted to another process? How
about converting and expanding in the same project? The answers dont come easily.
This presentation deals with issues to consider when making these decisions. Each
process has its advantages and disadvantages. Capital and operating costs for the
plant as a whole vary greatly depending on which process is selected. Entire plant
sections for grinding rock, handling of weak acid, or clarifying product acid are either
necessary or not, depending on the process. There major differences between
processes regarding size of facilities required to concentrate acid or to cool process
water. Operating costs vary greatly due to differences in energy efficiency, raw material
consumption, ability to consume water, and requirements for steam, cooling water, and
reagents.
Opportunities to recover valuable by-products are greatly affected by which process is
employed. Uranium recovery has recently re-emerged as an important issue, since
even environmentalists promote nuclear powers ecological advantages over fossil
fuels. Potential developments in uranium recovery from Hemi-Di plants might make this
even more attractive. The opposite is true for the Hemi process, which may be
incapable of practical uranium recovery. The Di process continues to be a proven
source for uranium recovery. The Hemi-Di process provides the highest quality gypsum
for a variety of gypsum utilization needs. The Di process can provide the greatest
recovery of fluosilicic acid.
This presentation highlights issues affecting capital and operating costs as well as byproduct recovery. The processes will be compared for new facilities and for potential
conversions of existing plants. A third area with quite different incentives is when an
existing facility might be expanded in capacity while converting to a new process.

#
John Wing, P.E., 1121 Waters Edge Drive, Lakeland FL 33801, USA
Phone: 1-863-666-6555
Email: johnwingpe@verizon.net

SELECTING A PHOSPHORIC ACID PROCESS

TABLE OF CONTENTS
PAGE
1

SUMMARY
TABLE OF CONTENTS

CHARACTERISTICS OF DI, HEMI, & HEMI-DI PROCESSES

ITS ALL IN THE GYPSUM
DIHYDRATE PROCESS
HEMI PROCESS
HEMI-DI PROCESS
OTHER PROCESSES
CALCIUM SULFATE CRYSTALLIZATION GRAPH

3
3
4
5
5
6

COMPARING PROCESSES
CAPITAL COST REACTOR/FILTER SYSTEM
CAPITAL COST OF OTHER PLANT SECTIONS
PHOSPHATE ROCK & SULFURIC ACID REQUIREMENT
ACID CONCENTRATION
ENERGY EFFICIENCY
AVOIDING ROCK GRINDING
OPERATOR, MAINTENANCE, AND CLEANING COST
OPERATING STABILITY
COOLING WATER REQUIREMENT
ACID STORAGE & CLARIFICATION
PHOSPHORIC ACID PURITY& Benefits to Fertilizer Analysis
REAGENT REQUIREMENTS
WATER CONSUMPTION INTO THE PROCESS
PROCESSING IMPURE ROCK
GYPSUM UTILIZATION
URANIUM RECOVERY
FLUOSILICIC ACID RECOVERY
EFFECT OF RECENT TRENDS

7
8
9
10
11
12
12
13
13
14
14
14
15
16
16
17
17
18

DI, HEMI, & HEMI-DI PROCESS COMPARISON TABLE

CONVERTING FROM DI TO HEMI OR HEMI-DI
CONVERTING TO HEMI OR HEMI-DI WHILE EXPANDING
REFERENCES
ABOUT THE AUTHOR

19
20
21
22
23

CHARACTERISTICS OF DI, HEMI, & HEMI-DI PROCESSES

ITS ALL IN THE GYPSUM
Phosphoric acid plants are really gypsum plants, because they make much more
gypsum than phosphoric acid. Since gypsum is often a bothersome waste product, the
plants are named for the valuable second product - phosphoric acid. However, the
various processes that are utilized are named for the type of gypsum that is produced.
Gypsum is calcium sulfate with various amounts of water of hydration attached to the
calcium sulfate molecules. The key to successful operation of a phosphoric acid plant is
to make good gypsum. Characteristics of good gypsum are large crystals that filter well
and a minimum of phosphate content within the gypsum crystals.
Gypsum crystals are in the Di (dihydrate) form at lower concentration and temperatures,
and in the Hemi (hemihydrate) form at higher concentration and temperature. The
Dihydrate (Di) process makes gypsum in the form of calcium sulfate dihydrate, which
has two water molecules per calcium sulfate molecule. The Hemi process makes
gypsum in the form of calcium sulfate hemihydrate, which has half a water molecule per
calcium sulfate molecule. The hemi and di zones are illustrated in the Calcium Sulfate
Crystallization Graph on page 6.
Either a hemihydrate or dihydrate process can have stable operation if the conditions
are clearly in either the hemihydrate or dihydrate zone. Problems occur between the
zones, because the crystals dont know which form they are supposed to be, resulting in
poor crystals and formation of scale. Think of the transition zone boundary as a line of
dragons that needs to be avoided where possible. Special techniques are necessary
where conditions must cross this transition zone, such as occurs on a Hemi filter. The
Hemi process became successful only after good dragon-fighting techniques were
developed for filtration and gypsum disposal.

DIHYDRATE PROCESS
This was the conventional process for most of the 20th century. Dihydrate plants have
made the phosphoric acid for most of the high analysis phosphate fertilizer that has ever
been produced. This process has a long track record or reliable operation, but it lacks
the energy efficiency and many of the operating advantages of the Hemi process. Most
phosphate rocks must be finely ground before processing.
Operating conditions in the Di process stay below the Hemi/Di transition boundary, but it
is economically necessary to push as deeply as practical into that boundary zone. The
filter product phosphoric acid is typically only 25-29% P205, so substantial further
concentration of product acid is required before making phosphate fertilizers.
Innovations have been used to expand capacity of some dihydrate plants to more than
double their original capacity.

Dihydrate process advantages include:

C
Long track record of experience
C
Predictable performance
C
High capacity relative to equipment size
C
Moderate recovery and sulfuric acid requirement
C
Proven potential for recovery of uranium by-product
C
Best for recovery of fluosilicic acid by-product
Disadvantages include:
C
Fine grinding of rock is normally required
C
Acid must be further concentrated to make most phosphate fertilizers.
C
Large steam and cooling water requirement
HEMI PROCESS
The Hemi (hemihydrate) process produces phosphoric acid directly from filtration at 4045% P205 concentration. Most Hemi plants use phosphate rock as received without
drying or grinding. Two entire plant sections are usually rendered unnecessary:
C
Evaporation to ~42% P205
C
Rock grinding (when using concentrate or other rock smaller than 2 mm)
Cooling water, acid storage, clarification, and steam distribution systems are reduced to
a fraction of their conventional size. Capital cost for the phosphate complex is roughly
20-25% less than for a dihydrate-based complex, which would require rock grinding,
evaporation, larger cooling water and steam distribution systems, and often elaborate
acid clarification systems.
Modern Hemi phosphoric acid plants tend to be easier to operate and require less
cleaning than dihydrate plants. One reason is that the reaction takes place in a stable
range of hemihydrate crystals. In contrast, dihydrate plants must (out of economic
necessity) operate near the unstable transition between dihydrate and hemihydrate.
Hemi process advantages include:
C
Minimum capital cost
C
Energy benefit from needing little or no steam to concentrate acid
C
Eliminate 27-42% evaporators
C
Usually eliminate rock grinding
C
Low cooling water requirement
C
Moderate phosphate recovery
C
Added recovery benefit where gypsum water is recirculated
C
Low sulfuric acid requirement
C
Easy to run and maintain; tolerant of process upset
C
Higher analysis fertilizer due to purer acid
Hemi has become the preferred process for making phosphoric acid in the 21st century.
Early Hemi plants were difficult to operate because of scaling problems that occurred
because of having to cross the zone of transition between Hemi and Di gypsum
crystals. During the last few decades people have developed ways to enjoy hemis high
concentration advantage without suffering its potential chaos.
4

HEMI-DI PROCESS
This advanced process begins with a Hemi reactor and Hemi filtration section, but it
adds a transformation reactor and a second filtration. The payoff for the added cost and
complication is extremely high recovery and high quality gypsum.
Hemi-Di advantages include:
C
98-99% P205 recovery
C
Very low sulfuric acid requirement
C
Energy benefit from needing little or no steam to concentrate acid
C
Eliminate 27-42% evaporators
C
Usually eliminate rock grinding
C
Low cooling water requirement
C
Gypsum purity is suitable for making a variety of by-products
C
Potential for enhanced uranium recovery (to be confirmed)
C
Higher analysis fertilizer

OTHER PROCESSES
Other phosphoric acid processes including Di-Hemi and a short-cut Hemi-Di have also
found their niche. If thats not enough, ask someone from Prayon to explain the HemiDi-Hemi process. Jim Hebbard can amuse you regarding the FIPR process, which
operates in the mono-calcium phosphate mode. There are other processes that dont
even make gypsum, because they dont use sulfuric acid, (so what I said earlier about
phosphoric acid plants being gypsum plant isnt always true). This presentation will deal
with only the three most common processes Di, Hemi, and Hemi-Di. I dont know
much about the other processes to cover them adequately, although I once convinced a
client not to pursue a Di-Hemi process.

Calcium Sulfate Crystallization Graph

C ourtesy of H ydro

COMPARING PROCESSES
Selection of the optimum process requires careful evaluation of capital and operating
costs, requirements for raw materials and reagents, quality of products and by-products,
utility situation, opportunities to recover valuable raw materials, and a variety of other
issues. Every project has its own set of needs, costs and opportunities, so a detailed
evaluation is necessary to pick the process. Factors which often vary so greatly that
they have overwhelming influence on process selection include:

Performance of phosphate rock in the various processes

Phosphate rock cost, delivered
Sulfur cost, delivered
Value of exported electric power
Gypsum utilization or disposal situation
Opportunity to recover uranium

The main issues are discussed below in a qualitative manner, based on typical or
average situations. Any project that is being seriously considered should involve a
comprehensive evaluation of each of these issues, including cost information that is
specifically tailored to the site.
To help provide an overview of issues to consider, the three processes are rated in the
attached Di, Hemi, and Hemi-Di Processes Comparison Table on page 19. This table
relates to average situations, but one must recognize that any individual project may
have circumstances that are far from typical.

CAPITAL COST REACTOR/FILTER SYSTEM

A Di process will usually have the least expensive main reactor. Experience with
dihydrate reactors over several decades has demonstrated that many plants are
operating at capacities that are more than double original design capacity. Such
experience is utilized to confirm that new reactors can be much smaller than those of a
few decades ago. Hemi reactor size requirement has also tended to decrease as plant
experience is gained, but the Hemi process has a shorter track record to draw from.
Filter size requirement varies greatly with the type of phosphate rock that is utilized.
Pilot plant testing of the phosphate rock can predict filtration rate before the plant is
designed. It has been my observation that filtration rate tends to average somewhat
higher with a Di process than with a Hemi process. However, there are many
exceptions, so a valid comparison between the two processes must consider
experience in real plants and pilot plant testing.
Since the Hemi process operates at hotter temperatures, metals tend to corrode
somewhat faster than in the Di process. Consequently, metals in agitators, pumps, and
filter wetted parts will usually be somewhat more expensive in a Hemi plant. Flash
coolers in Hemi reactors can be significantly smaller and require less circulation rate
7

than those in Di plants. However, Hemi plants require more fume collection capacity
and efficiency than Di plants.
The entire reactor/filter system for a dihydrate process will usually be smaller and less
expensive than for a Hemi process. A Hemi-Di process requires a Hemi reactor/filter
system followed by a simpler and smaller dihydrate reactor/filter system. Consequently,
reactor/filter capital cost is very high in a Hemi-Di plant.

CAPITAL COST OF OTHER PLANT SECTIONS

The Hemi process has advantages outside of the reaction and filtration sections that
tend to allow the total plant capital cost to be significantly less than with a Di process.
The first advantage is that the Hemi reaction does not require the phosphate rock to be
ground as finely as does the Di process. The popular Yara (Hydro) Hemi process can
use rock particles as large as 2 mm (9 mesh), compared to a typical 35 mesh
requirement with the Di process. This means that a Hemi plant can use phosphate
concentrate and other phosphate up to 2 mm in size without any processing. In
contrast, a Di process usually requires rock to be finely ground in a ball mill, requiring
major expense for capital, operation, maintenance, and energy.
A Di plant makes filter product acid at around 27% P205 concentration, compared to
about 42% P205 acid from Hemi filters. Consequently it requires additional evaporators
to concentrate this acid, along with substantial investment for additional cooling water
and acid storage facilities.
A Hemi plant normally has lowest capital cost for a new plant, because it usually avoids
rock grinding (except with pebble-sized rock), it requires much less (if any) evaporation
of acid, cooling water is minimized, there is no 27% P205 acid to handle, and product
acid clarification is usually unnecessary.
Although a Di process might have a less expensive reactor and filter section, it would
require all of those extra plant sections and expanded capacities. Consequently, the
whole plant containing the Di process would normally cost around 20-25% more than a
Hemi process plant.
A Hemi-Di process shares Hemis advantages of eliminating or down-sizing of those
plant sections. However, Hemi-Di requires a second (but simpler) set of reaction and
filtration facilities, which roughly offset savings in other sections. I would guess that the
whole Hemi-Di plant would cost no more than a whole Di plant, when cost of rock
grinding, evaporators, acid handling, and cooling are considered.

PHOSPHATE ROCK & SULFURIC ACID REQUIREMENT

Most loss of product in a phosphoric acid is in the filter cake, so filter recovery has the
largest influence on how much raw materials are required. This must be evaluated on a
case by case basis. Licensors guarantee performance including recovery at the filter
cake for the specific plant and conditions. Di and Hemi plants typically achieve filter
recovery in the mid-90% range, with Di plants averaging perhaps a percent or so higher
than Hemi plants. Hemi-Di recovery (based on filter cake) is typically 98-99%.
There will be considerable additional losses beyond filter cake losses. Phosphoric
acid is lost by entrainment from flash coolers and evaporators, leakage from seals, filter
mechanical losses, dumping of acid during maintenance and cleaning, and accidental
spills. The total of such losses may amount to a percent or two of product in a typical
plant. Hemi & Hemi-Di plants have much less evaporator entrainment losses, because
very little evaporation of acid is required. Only Di plants are exposed to losses from
handling 27% P205 acid because Hemi & Hemi-Di dont produce the weak acid. Hemi-Di
plants have additional exposure to general material handling losses around reactors and
filters, because they have more reactors and filters.
There can be opportunities to recover what had been lost. Many plants recirculate the
acidic water from the gypsum stacks and/or cooling water systems. Some of this acidic
water is used to wash the filter cake. Most cake wash water proceeds through the filter
and ends up in product acid. Thereby, over half of all phosphoric acid which had been
lost into this acidic water is eventually recovered from the recirculated water.
Any plant that recirculates acidic water from the gypsum stack will recover most of the
water soluble losses in the filter cake. However, water insoluble losses from a Di plant
(or the Di filter in a Hemi-Di plant) remain locked in the dihydrate gypsum crystals, so it
cannot get recovered. This refers to losses that are reported as CI losses (citrate
insoluble un-dissolved rock) and CS losses (citrate soluble co-precipitated dicalcium
phosphate in gypsum crystals the majority of filter losses).
A Hemi plant that recirculates acidic water from the gypsum stack has an added
recovery benefit. This is because the hemihydrate gypsum crystals will release coprecipitated dicalcium phosphate out of the crystals and into the surrounding water.
will recover most of the water soluble losses from the filter cake. For example, a
typical Hemi filter cake contains co-precipitated dicalcium phosphate (citrate soluble
P205) equivalent to 3.5% of the feed. After discharging from the filter the hemihydrate
crystals gradually dissolve while forming dihydrate crystals. The co-precipitated
dicalcium phosphate dissolves in the acidic water, and very little re-precipitates into the
dihydrate crystals. Thus about half of that 3.5% loss (1.75%) gets recovered when
recirculated water is used to wash the filter cake.

P205 Recovery Rankings for the three processes are typically as follows. High recovery
means low phosphate rock requirement.
1st:

HEMI-DI
High recovery 98-99% is the reason for this fancy process

2nd:

HEMI with Recirculation of Water from Gypsum Stack

Recovery of an extra couple percent of lost P205 gives Hemi the
advantage over Di.

3rd:

TIE: HEMI & DI

Di usually has a slight edge with recovery based on filter cake, but Hemi
avoids evaporator entrainment and weak acid handling losses.

A Hemi reactor needs significantly less sulfuric acid than a Di reactor, primarily because
of the reduced ratio of sulfate to phosphate in the reactor acid. Acid from Hemi reactors
typically contains 2% free SO4 and 43% P205 (a ratio of 0.46), whereas Di reactor acid is
typically 2% free SO4 and 27% P205 (a ratio of 0.74). Another small advantage with
Hemi is that slightly less sulfuric acid reacts with aluminum impurities. In the Yara
(Hydro) Hemi process these effects amount to needing about 2.5% less sulfuric acid
than with a Di process with similar recovery.
Sulfuric acid consumption will track with P205 recovery, except that Hemi & Hemi-Di
have about 2.5% advantage due primarily to low-sulfate product. The processes
typically rank in this order regarding sulfuric acid consumption: Hemi-Di, Hemi, and Di.

ACID CONCENTRATION
Most dihydrate phos acid plants make 25-29% P2O5 product. Higher product
concentration is impractical, because it would involve pushing operating conditions into
the unstable hemi/di transition boundary.
Hemi plants produce phosphoric acid directly from filtration at concentrations between
38% and 46% P2O5. Optimum concentration has been around 43% P2O5, which is near
the sweet spot where a Hemi plant performs best. Rapidly increasing energy value
may entice anyone who operates or designs a Hemi plant to raise product concentration
to further enhance energy efficiency. Higher product concentration would require
somewhat larger reactor and filters, and recovery might decline. However, these tradeoffs may be wise, considering the energy benefits to be gained.
Hemi-Di plants benefit from water balance and cake washing situations that make it
practical to make even higher concentrations of phosphoric acid ranging from 40% to
50% P2O5.

10

ENERGY EFFICIENCY
Energy efficiency is essential in dealing with global warming. The Hemi process forte is
efficient use of energy. Its high product acid concentration avoids the need for the huge
quantity of evaporator steam that would otherwise be required to make product
concentration suitable for further processing. Ongoing trends of soaring energy costs
and need to conserve energy will further magnify this advantage in coming years.
The biggest source of energy for a typical phosphate chemical complex is sulfuric acid
production. Surplus heat from burning of sulfur is absorbed by steam, which is used to
generate all electric power required by the complex plus an export of power.
Energy efficiency of a phosphate complex is greatly enhanced by use of the Hemi
phosphoric acid process. A 1,500 T/D P2O5 phosphoric acid plant will save about 2500
T/D in evaporator steam by making filtered acid at 42% P2O5, compared to dihydrate
process acid at 26% P2O5. This surplus steam would typically be used to generate
electric power. Total electric power production is near 60 megawatts for a 1500 metric
T/D P2O5 phosphate complex. This is worth $35 to $50 million annually, based on
electric power values of 7 to 10 cents/kwh. The energy advantages of the hemihydrate
process account for about 16 megawatts of this power worth $9-15 million/year or
$18-30/ton of P2O5. The surplus electric power could be exported to the power grid for
sale, or it could be wheeled to the owners mine or other nearby facilities. Future
electric power values are likely to increase substantially as other sources of energy
become increasingly expensive.
A 1500 T/D Hemi or Hemi-Di plant would typically save another 3 MWs of electric power
by not having to grind rock, and by having much smaller acid evaporation requirement.
Hemi & Hemi Di plants also save the electric motor power that Di plants need to operate
additional evaporators, cooling water pumping, and acid handling facilities. Hemi-Di
plants do require substantial power for the second reactor and filters, thus offsetting part
of the motor power savings.
It is important to note that any utilization of energy from waste heat is environmentally
friendly. This electric power is produced with incremental net results of no pollution, no
greenhouse gas, no solid waste, and no consumption of fuel. No other source of
energy can top this for ecological responsibility - whether it uses coal, oil, gas,
nuclear fuel, wind, or solar energy.
Energy Efficiency rankings for the three processes would be:
1st:

HEMI

Needs little or no evaporator steam

Avoids rock grinding (usually)
Less power for evaporation, acid handling, and cooling water

2nd:

HEMI-DI

Same advantages as Hemi

However, the second reactor and filters need more power.

3rd:

DI

A distant last place.

11

AVOIDING ROCK GRINDING

The Yara (Hydro) Hemi phosphoric acid process can use rock which is much coarser
than that required for conventional dihydrate processes, so grinding is not required for
most of the worlds phosphate rock sources. Particle size requirement for Yara Hemi is
typically -9 mesh (-2 mm), versus -35 mesh (0.42 mm) for typical dihydrate processes.
The Hemi process can handle damp rock with up to about 15-20% moisture. Most
commercial phosphate rock sources worldwide are suitable in particle size and moisture
content for feeding directly to a Hemi plant without drying or grinding. This includes
coarse concentrate and some screened phosphate rocks.
An important exception is that the pebble rock which makes up a significant portion of
central Florida phosphate is much too coarse to feed directly to any phos acid plant.
Pebble rock could be ground to -2 mm size in relatively low energy roller, impact, or
hammer, mills with closed circuit screening. This pebble rock is available dripping wet
with about 10% moisture, and drying would be quite expensive. Consequently, the
recommended method would be to grind the damp rock without drying, followed by wet
screening and recycle of damp +2 mm material to the mill. Such milling requires only a
fraction of the power and capital cost that a ball mill requires.

OPERATOR, MAINTENANCE, AND CLEANING COST

Both Hemi and Hemi-Di benefit from elimination of rock mills and evaporators, and have
smaller acid handling and cooling water sections than a Di plant. Hemi reactors and
flash coolers stay cleaner than those in Di plants. Di filters stay cleaner than Hemi
filters. A Hemi-Di process has a second reaction and filtration second to operate and
maintain. A Hemi plant should have a significant advantage over either Di or Hemi-Di,
because there is much less equipment to operate and maintain.
Hemi reactors have a major stability advantage, because they operate in a stable zone,
well above the hemi/di transition boundary. Substantial changes in temperature and
concentration can be tolerated without approaching the transition. Operating control is
less critical, and the reactor is more forgiving to upset conditions or sudden changes in
rock feed characteristics. This accounts for praise by those that operate the plants that
they are easier to operate, more stable, and more forgiving than dihydrate plants.
There is relatively little scale formation in the reactors and flash coolers, because of
operating in the stable zone, and because there is lower solubility of calcium sulfate.
Hotter conditions in Hemi reactors and filters cause faster corrosion to agitators, filter
metal surfaces, etc. Optimum metals for Hemi service are typically one step up from
metals that would be optimum for dihydrate service. Hemi reactor agitators are typically
in the 904L or Ferallium 255 class, although existing 317L agitators have lasted fairly
well in plants that were retrofitted from di to hemi. Upper agitator shafts in either hemi
or di reactors require rubber coating. Filter pans are normally 317L or either process,
but Hemi plants with highly corrosive acids should use 904L. Belt filter vacuum boxes
are typically 904L.
12

OPERATING STABILITY
Either a hemihydrate or dihydrate process can operate stably if the conditions are
clearly in either the hemihydrate or dihydrate zone. The hemi/di transition zone is
illustrated in the Calcium Sulfate Crystallization Graph on page 6.
Dihydrate plants must limit reactor concentration in order to keep below a transition
zone between hemihydrate and dihydrate. When concentration or temperature gets a
little too high, the gypsum crystals form as a mixture of dihydrate and hemihydrate
crystals. These crystals are small, which reduces filtration rate. Wherever the slurry
cools, scale forms inside the reactor, pumps and piping. For economic reasons
dihydrate plants must push slightly into the transition zone, but good control can
minimize problems. If a dihydrate reactor is allowed to get seriously over optimum
temperature or concentration, filtration becomes extremely slow, and equipment scaling
is severe. Crystals in a typical dihydrate plant are a mixture of some hemihydrate
among mostly dihydrate crystals. Dihydrate plants tend to have far more scale formation
in reaction and flash cooling systems than hemi plants.
Hemi reactors have a major advantage, because they operate in a stable zone, well
above the hemi/di transition zone.
Substantial changes in temperature and
concentration can be tolerated without getting into the transition zone. Operating
control is less critical, and the reactor is more forgiving to upset conditions or sudden
changes in rock feed characteristics. This accounts for praise by those that operate the
plants that they are easier to operate, more stable, and more forgiving than dihydrate
plants. There is relatively little scale formation in the reactors and flash coolers,
because of operating in the stable zone, and because there is lower solubility of calcium
sulfate.
The Calcium Sulfate Crystallization Graph shows a hemi/anhydrite transition above
the hemi zone. In actual practice this transition is so high that it is rarely a problem,
except occasionally in some hemi-di plants which push reactor acid concentration to 4850% P205.
In a Hemi plant conditions in filtration pass thru the Hemihydrate/Dihydrate transition
line. Crossing this transition caused problems for early Hemi plants. However,
technology has been developed which allows this transition to be crossed with minimal
scaling in the filter system. Part of this technology has been use of an anti-scalant
reagent that greatly slows the conversion of hemihydrate crystals to dihydrate crystals,
thus reducing scaling. Anti-scalant is not always necessary, as was demonstrated at
Belledune, where the anti-scalant system was abandoned.

COOLING WATER REQUIREMENT

Hemi & Hemi-Di plants have less need for the huge flow of cooling water normally
required by phos acid evaporator condensers - perhaps none at all. A relatively small
flow of cooling water is required for flash cooler condensers, fume scrubbing, equipment
washing, etc. This water does not need to be as cool as the 33-350C (92-950F) required
13

for 52-54% P205 evaporators in a dihydrate phos acid plant. Water at around 38oC
(1000F) will be adequate for scrubber water, equipment wash, etc., and more than
adequate for flash cooler condenser water. Hemi & Hemi-Di plant cooling pond and
associated pumps and piping are a fraction the size of comparable equipment in a Di
plant.

ACID STORAGE AND CLARIFICATION

Phosphoric acid storage and clarification facilities in a Hemi or Hemi-Di plant are about
one third that required for a dihydrate plant. There will be agitated storage tanks for
42% acid (and possibly some 30% acid for DAP scrubber feed). Clarification is not
necessary because the 30% and 42% phosphoric acids have both come from filtration,
and because the high purity of Hemi acid makes further clarification unnecessary for
product analysis purposes.
A Di plant would typically require storage and clarification for 26% and 42% acid, plus
agitated storage (and often clarification) for 52-54% acid.

PHOSPHORIC ACID PURITY & BENEFITS TO FERTILIZER ANALYSIS

Phosphoric acid from a Hemi or Hemi-Di plant is purer than that from a Di process, with
lower sulfate, aluminum, fluoride, and solids content. DAP, MAP, and TSP produced
from this acid will be about 2 percentage points higher in P2O5 than that from a
dihydrate plant. This facilitates production of on-grade DAP and TSP from phosphate
with high impurity levels. MAP grade would rise, thus reducing shipping cost.
Effect of acid purity on DAP nitrogen content is more difficult to predict, because %N is
affected not only by product purity, but also by efficiency of ammonia absorption.
Calculated DAP grade benefit is over 0.5 percentage points in N, but this assumes
adequate ammonia absorption.
Effect of Hemi acid purity on DAP grade was demonstrated at the Belledune plant,
which was converted to Hemi in 1986, using 66-67 BPL central Florida rock. Before the
Hemi conversion is was difficult to meet 18-46-0 DAP grade, using settled 40% P205
acid. Upon Hemi start-up, grade jumped to about 47.5% P205 and easily exceeded the
18% N requirement, using un-settled 40% acid. A simple modification was employed to
prevent over-formulation of DAP.

REAGENT REQUIREMENTS
Nearly all phosphoric acid reactors require defoamer, and I have not noticed significant
difference between Di and Hemi reactors. However, Arcadian/PCS found that when
using BuCraa rock from the Western Sahara region of Morocco, no defoamer was
required with the Di process or after conversion to the Hemi process.

14

Yara prescribes a proprietary anti-scalant to minimize formation of scale within filters in

a Hemi plant (but not in a Hemi-Di plant). This costs roughly $1/ton P205. An exception
was that the Belledune Hemi plant discontinued use of the anti-scalant and removed the
system. That was with a Bird tilting pan filter, using phosphate rock from central Florida
and Morocco.
Hemi and Hemi-Di plants sometimes add clay to the hemi reactor to modify crystal
shape and/or absorb corrosive free HF.
Hemi-Di plants sometimes add clay to the second reactor to facilitate transformation of
hemi gypsum crystals to dihydrate form.
WATER CONSUMPTION INTO THE PROCESS
A Di process must consume much more water into the process than a Hemi or Hemi-Di
process, largely due to the effect on water balance for producing 27% versus 42% P205
product. Water consumption can be good, bad, or of little consequence, depending on
the circumstance. A plant receiving ground rock slurry from a wet ball mill would have to
have a Di process. That much water would drown a Hemi reactor. However, a Hemi
reactor does not need finely ground rock, so a wet ball mill should not be involved
anyway.
There may be other reasons to want to consume extra water into a reactor. For
example, one client expressed interest in a Hemi conversion, but the plant received
70% sulfuric acid at very low cost. There was too much water in 70% H2SO4 for the
Hemi process to work, so the plant kept its Di process. The Di process could have an
advantage where phosphate rock is received as a slurry by pipeline, because it might
be able to consume the entire slurry without de-watering. Such rock slurry would have
to be partly de-watered to feed a Hemi plant.
For plants in a desert or other location where water is very expensive, the Hemi
reactors lesser need for water is an advantage. The Hemi plant also has an advantage
at the gypsum discharge end. Free moisture in filter cake is about the same for a Di or
Hemi filter. However, Hemi filter cake contains far less water of hydration, because of
the difference between calcium sulfate hemihydrate versus calcium sulfate dihydrate.
Plants in deserts often discharge gypsum to the gypsum stack as filter cake, rather than
as slurry. In such plants the difference in total water content in the gypsum is important.
Hemi gypsum cake is self drying because hemi gypsum absorbs most of the free
water from the filter cake as gypsum converts from hemi form to dihydrate form. In
desert conditions, this leads to another advantage of discharging Hemi filter cake. This
self drying gypsum will never seep free water downward. Thus there is no real need
for using an expensive water-proof liner under a hemi gypsum stack. In some situations
it might be necessary for the plant owner to convince regulatory authorities not to
mandate a multi-million dollar expense for a useless liner.
The same self drying gypsum advantage is achieved with a Di-Hemi or Hemi-Di-Hemi
process. It is not achieved with a Hemi-Di process, because Hemi-Di discharges
dihydrate gypsum from the Di filters.
15

PROCESSING IMPURE ROCK

There is increasing evidence that the Hemi process performs well with some types of
phosphate rock that are impractical to process with conventional dihydrate processes.
This includes rock with very low P205 concentrations and unusually high levels of iron,
aluminum, and magnesium impurities. Such rock is either left in the ground, blended
with higher grade rock, or processed for further purification. Ability to process low-grade
rock could lead to breakthroughs in utilizing phosphate reserves in places like South
Florida.
One Hemi plant successfully uses a large amount of sub-commercial rock that had
been considered practically worthless for processing. A company that has both Hemi
and Di plants routinely routes the lowest grade rock to the Hemi plant, because it
processes it more easily. At Belledune the original Di plant was unable to make ongrade DAP from clarified acid; when the plant was converted to Hemi, it immediately
exceeded DAP grade by a wide margin.
GYPSUM UTILIZATION
Gypsum from phosphoric acid plants can be utilized to produce a variety of products.
Such phospho-gypsum has been used for road-bed material, cement ingredient, sulfate
fertilizer, agricultural soil conditioner, or gypsum wall board. Further processing of the
gypsum can produce limestone, ammonium sulfate, sulfuric acid, glass, and ceramics.
The Florida Institute for Phosphate Research (www.fipr.state.fl.us) is heavily involved
with an international effort to develop many uses for billions of tons of phospho-gypsum.
FIPR has stated that research indicates that a beneficial, commercially appropriate and
environmental neutral use for PG (phospho-gypsum) would be preferable to dumping or
perpetual storage in stacks. This effort is supported by international and regional
phosphate organizations including IFA, IMPHOS, AFA, and Mosaic, as well as the
OECD, IAEA, and other organizations.
In the United States it is unfortunate that the EPA banned use of phospho-gypsum in
1992 essentially mandating perpetual storage for billions of tons of this potentially
valuable material. This resistance seems to be an overreaction to the minor radioactive
content of phospho-gypsum. Although original flawed data that led to this conclusion
has been repudiated, efforts to have the EPA re-evaluate their stance have proceeded
extremely slowly.
Processes that utilize gypsum from phosphoric acid plants often require the gypsum to
have limited concentration of phosphate and fluoride impurities. Suitable high purity
phospho-gypsum can be produced from either a Hemi-Di process or a Di-Hemi process.
Gypsum of moderate purity is obtained by re-washing gypsum cake from phosphoric
acid filters.
Hemi, Di-Hemi, and Hemi-Di-Hemi processes produce gypsum filter cake that is selfdrying, which can be a major advantage for drying requirements when gypsum is
utilized. This self-drying occurs because hemihydrate gypsum absorbs most of the free
water from filter cake when it converts to dihydrate gypsum.
16

URANIUM RECOVERY
Phosphate rock contains uranium in concentrations up to a pound (half a kilogram) of
U3O8 per ton of rock. Uranium can be recovered by solvent extraction from phosphoric
acid from dihydrate plants containing about 25-29% P205. This extraction is preceded
by extensive clarification and pre-treatment of the phosphoric acid. Much uranium was
recovered from phosphoric acid a couple of decades ago. A sudden downturn in
uranium price forced recovery plants to shut down.
Now nuclear power plants are finding favor as economically and environmentally
attractive sources of electric power. Despite concerns about disposal of spent fuel,
nuclear power plants are often considered a more environmentally friendly power
source than fossil fuels. They emit no air or water pollution and no greenhouse gasses.
The Nuclear Regulatory Commission is handling 16 applications for nuclear power
plants in the USA, and expects to receive 32 by 2012.
Uranium prices have soared to double or triple what they were a couple decades ago.
Even so, uranium costs only a fraction as much as oil per unit of power produced.
The Di process has a proven track record of successful uranium recovery. The Hemi
process is not attractive for uranium recovery. Its high product acid concentration makes
uranium extraction difficult, and uranium content in the acid is low.
Uranium recovery might be exceptionally attractive with a Hemi-Di process. This is
because a high uranium-to-P205 ratio occurs in a certain weak acid filtrate stream with
the Hemi-Di process. It appears that it would be far easier to extract uranium from this
stream than with the conventional extraction from 27% P205 acid, and quantity of
uranium might increase. However, this process has yet to be proven and developed.

FLUOSILICIC ACID RECOVERY

Fluosilicic acid (FSA) is sometimes recovered from phosphoric acid plants. When it
contains relatively high concentrations of P205 (0.1-0.25%), it is sold at relatively low
price for use in fluoridating municipal water. When P205 is limited to a few hundred ppm,
it is sold at a higher price for making aluminum fluoride.
The Di process excels in opportunity to recover FSA, because FSA can be recovered
from the evaporators that concentrate acid from 27-42% acid, as well as from the higher
concentration evaporators. Hemi and Hemi-Di processes eliminate the need to
evaporate acid from 27 to 42% P205, so this opportunity to recover FSA from 27-42%
evaporators does not exist. However, some FSA is recovered from Hemi reactor fumes
or from flash cooler vapors, as well as from evaporators for higher acid concentrations.

17

EFFECT OF RECENT TRENDS

Some recent trends will have a major impact on process selection. Soaring fuel costs
will obviously place major emphasis on the energy-efficient Hemi and Hemi-Di
processes. An added benefit of any such means of recovering waste heat is that it will
help combat global warming.
Recent sharp increases in cost of sulfur and phosphate rock make it more attractive to
invest in the added recovery benefit of the Hemi-Di process.
A flurry of interest in new nuclear power has escalated the value of uranium. The Di
process and probably the Hemi-Di process produce acid from which uranium can be
extracted economically.
If and when the US EPA gets around to relinquishing its ban on utilization of gypsum,
various uses of this valuable by-product will be pursued. The Hemi-Di process produces
gypsum with the high purity that is essential for some of these uses.

18

DI, HEMI, & HEMI-DI PROCESS COMPARISON TABLE

(Ratings with 5 being excellent)
CRITERIA

DI

HEMI

HEMI
-DI

Capital Cost, Reactor & Filters

Capital Cost, Other Sections

5
2

4
5

2
5

Operating Cost & Benefits

Energy Efficiency, Total Plant
Product Acid Concentration
Evaporators & Steam Reqt
P2O5 Recovery at Filter
P2O5 Losses Other Than Filter

2
1
1
1
3
3

4
5
4
4
2
4

5
4
5
4
5
3

Recovery of Losses from

Recirculated Water
Sulfuric Acid Consumption
Rock Size Requirement

3
1

3
4

5
4

Cooling Water Requirement

Product Clarification, Storage
Product Acid Purity

2
1
2

4
4
4

4
4
4

Capacity per Size of Eqip.

Reagent Requirements

4
4

3
3

2
3

Familiarity & Experience

Complexity of Operation

5
2

4
4

3
2

Uranium Recovery
Gypsum Utilization
Fluosilicic Acid Recovery

4
2
4

0
2
2

5?
4
2

19

REMARKS
(for typical or average situations, with exceptions)
Di has smallest Reac. & Filt. H-D has 2 Rx & Filter stages
Hemi & H-D need no rock grinding, less evaporation,
acid storage, & cooling water
Di needs rock grinding & much evap. H-D: high recovery
Di needs rock grinding, much more steam & cooling water
Di 25-29% P2O5, Hemi 40-45%, H-D 40-50%
Hemi & Hemi-Di make DAP with little or no evaporation.
Di ~96%, Hemi ~95%, Hemi-Di ~98.5%
Hemi & H-D avoid handling 27% acid.
H-D has 2nd Rx & Filter
Works only where water recirculates from gypsum stack.
2% benefit to Hemi & H-D due to low SO4 in product, etc.
Di needs <0.4 mm (35 mesh).
Hemi & H-D can use <2 mm (9 mesh)
Hemi & H-D need no 42% evap. condenser water
Hemi & H-D have no 27% acid; often need no clarification
Hemi & H-D acid can make fertilizers with 2% more
%P2 O5
Di has smallest equip. H-D requires 2nd reactor & filters
Hemi may need anti-scalant
Hemi & Hemi-Di may use clay or silica.
Most existing plants are Di, but many are Hemi & H-D.
Di needs grinding, much evaporation, etc.
Hemi-Di has 2nd Reactor & Filters
Hemi-Di may be best, but needs development.
Hemi-Di gypsum is purest.
Hemi & H-D have no 27-42% evaps, hence less FSA.

CONVERTING FROM DI TO HEMI OR HEMI-DI

When should one consider converting an existing Di plant to Hemi or Hemi-Di? A key
issue is that the Hemi process requires very little evaporation of product acid. The huge
quantity of steam that had been going to the evaporators becomes available, so the
decision is largely based on how much value can be obtained from all of that steam.
The surplus steam would normally be used to generate electric power. If there is
surplus capacity in an existing power co-generation facility, and if this power can be
used effectively or sold at a good price, then there is major justification for converting
the plant to the Hemi process. When electric power was cheaper, it was difficult to
justify the expense of new power co-generation facilities. Now electricity is so valuable
that this old rule of thumb no longer holds true.
If a Di plant needs another evaporator, one should consider converting to Hemi instead
of buying the evaporator. Since a Hemi plant avoids the need to concentrate acid from
27 to 42% P205, conversion to Hemi would eliminate any shortage of evaporation
capacity. The capital that is saved by avoiding one new evaporator may cover most of
the cost of a Hemi conversion. Furthermore, a Hemi conversion would greatly reduce
need for steam, cooling water, and acid storage facilities potentially bringing additional
capital cost savings.
If a Di plant is having difficulty meeting grade with DAP or TSP, a Hemi conversion
would increase P205 content in DAP, TSP, or MAP by 2 percentage points. It should
also help N concentration, depending on ability of the product to consume ammonia.
Where there are two or more phos acid plants, converting only one of them to Hemi
may solve DAP or TSP grade problems for the entire facility.
The Belledune Fertilizer plant in New Brunswick Canada is an example of a very
profitable Hemi conversion. The plant might have been shut down because of high cost
and difficulty in meeting DAP grade. After converting an old Prayon Mark 2 dihydrate to
Hemi in 1986, cost were slashed by totally eliminating the evaporation section and
associated fuel cost for generating steam. Recovery averaged 95%, and capacity
easily topped the modest increase in design rate limited only by raw material and
product requirements. The superintendent called it one sweet plant to run. Belledune
continued to operate for a decade, and was considered one of the worlds easiest
running phosphoric acid plants. DAP grade became easy to reach, using 40% acid that
needed no settling to remove solids.
Further conversion to Hemi-Di involves major expense for the second reaction and
filtration facilities. Justification for this expense comes from the major reduction in raw
material cost that is achieved by Hemi-Dis 98-99% recovery efficiency. Recent
increases in phosphate rock and sulfur prices make Hemi-Di especially attractive.

20

CONVERTING TO HEMI OR HEMI-DI WHILE EXPANDING

Additional economic opportunities arise when simultaneously converting an existing
dihydrate plant to Hemi or Hemi-Di while expanding capacity. First, a major expansion
can be made without adding evaporators, because Hemi needs so little evaporation.
Second, the existing cooling water system will accommodate a major expansion,
because of savings in evaporator condenser cooling water requirements. Third, the
rock grinding section is likely to be eliminated, so no expansion is required there.
The acid storage tank area may not need expansion when a Di plant is converted to
Hemi or Hemi-Di of substantially greater capacity. Tanks that had been used for 27%
acid storage will become available for other acid storage service. Acid clarification
requirements are reduced or eliminated, because the Hemi acid will be purer.
Arcadian (now PCS) in Louisiana made good use of those down-stream advantages
when they expanded capacity by a third while converting to the Hydro Hemi process.
After hearing of Belledunes success, Arcadian converted its Prayon Mark 2 plant with
Bird filter to Hemi. Expenses beyond the reactor and filter sections were minimized
because:
C
Rock Grinding was totally abandoned and by-passed, with un-ground BuCraa
rock feeding from a rock washing filter directly to the reactor.
C
Elimination of the requirement to concentrate acid from 27 to 42% P205 allowed
existing evaporators to make more capacity.
C
One evaporator was dedicated to boosting concentration from 54% to 60-62%
P205 for feeding a super-phosphoric acid facility, thus increasing super-acid rate.
C
Requirements for cooling water were reduced.
C
No new phos acid storage facilities were required.
The Arcadian Hemi plant started very easily achieving design capacity and conditions
within two days. The plant easily performed so well that the client accepted it without
doing the customary performance test run. It frequently ran at 110% of design capacity,
and occasionally achieved up to 130% of design capacity. Recovery consistently
exceeded 96%. Arcadian and Belledunes experience demonstrated that Prayon
reactors and tilting pan filters are well suited to conversion to the Hemi process.
The exceptionally high quality Hemi acid was welcome as feedstock to a food-grade
phosphoric acid plant and a super-phosphoric acid facility. Arcadians liquid fertilizer
product was considered to be the best in the domestic industry.
Expansion while converting a Di or Hemi plant to Hemi-Di can be facilitated by using the
existing filters in the dihydrate section of the Hemi-Di process. Dihydrate filtration in a
Hemi-Di plant needs only about 60% of the filter area as hemi filtration. The old filters
may be big enough to act as dihydrate filters in a new and larger Hemi-Di plant.
16Ap08 C:\Documents and Settings\John Wing\My Documents\P.E\Genl\Wing-CW08-Select PA Process.wpd

21

REFERENCES
Pierre Becker, Phosphates & Phosphoric Acid, 2nd Ed., Marcel Dekker Inc., NYC, 89.
BuShea, et al., Application of BSF Technologies CTC3 to a Phosphate Complex,
AIChE Natl Convention, Orlando, 3/90 and 1990 AIChE Clearwater Convention
J. David Crerar & Barry T. Crozier, Practical Retrofitting to the Hemihydrate Process
and Plant Performance Data, AIChE Meeting, Lakeland FL, Mar. 87.
B.T. Crozier, Fisons Hemihydrate Process - A Decade of Energy Savings, The
Fertilizer Institute Round Table, Atlanta, Oct.82.
John Gobbitt, Hemihydrate Phosphoric Acid Plant Retrofits at Geismar and Chinhae,
AIChE Clearwater Convention, May, 90.
Joseph W. Guida, Phosphoric Acid and Uranium Recovery - Take 3"
Fertilizer International, Jan-Feb., 2008
G.W. Hartman & L.J. Friedman, The Royster Power Program - Start-up and Initial
Operating Experience AIChE Convention, New Orleans LA, April,86.
Sam Houghtaling & John Wing, Hemi or Hemi-Di - Our Future, AIChE Spring National
Meeting, New Orleans LA, April,92.
Sam Houghtaling & John Wing Hemi or Hemi-Di? - Arcadian Converts Phos Acid Plant
from Dihydrate to Hemi, AIChE Clearwater Convention, May,91.
John Wing, The Hemi Era in Phosphoric Acid, AIChE Clearwater Convention, June,06
John Wing, From Phosphate Rock to DAP at Lower Cost, AIChE Clearwater
Convention, May,99.
John Wing, The Case for Converting Phos Acid Plants to Hemi, AIChE Clearwater
Convention, May,95.
John Wing, Hemihydrate Phosphoric Acid Plant Conversion at Belledune, Canada,
AIChE Meeting, Lakeland FL, Oct.,87.
John Wing, Florida Phosphate Technology - 2000", AIChE Clearwater Conv., May,89.
Anonymous, Hemi Forever? Fertilizer International, July-Aug. 2007, pages 43-48.
Anonymous, HiTech Solutions, Inc. - Contractor Profile. Phosphorus & Potassium
magazine, Jan.-Feb. 93, page 26.

22

ABOUT THE AUTHOR

John Wing - President of the consulting firm John W ing, P.E. - has a Bachelor
of Chemical Engineering degree with Honors from University of Florida and
a Master of Engineering in Administration from University of South Florida.
He has served the phosphate industry for decades in process design,
consulting, project management, technical service, process development,
and production supervision. He was vice president of HiTech Solutions and
was employed in other engineering and phosphate production companies.
He did process engineering for new plants, modifications and expansions
for many phosphoric acid, DAP, MAP, & TSP plants. He designed
fluosilicic acid recovery systems at Conserv in Florida, Sterlite and
Hindalco in India, and IJC in Jordan. He provided conceptual design for six
phosphoric acid evaporators and several dozen fume scrubbers.
For Hemi plants he performed the process design for the Belledune Hemi
conversion, the 3-train Maaden plant which is in progress, and conversion
of a very large dihydrate plant to Hemi-Di. He was project manager and
process engineer for the Arcadian Hemi conversion. He consulted as
owners engineer for the Indo-Jordan Chemical Co. Hemi plant in Jordan
and has provided subsequent technical services for the same plant.
He has written technical papers on:
Hemi and Hemi-Di processes (7 papers)
The future of the phosphate industry
Phosphoric acid evaporation
Cooling pond systems
Can a Little Altruism Enhance an Engineers Career Satisfaction?
Article segments for Phosphorus & Potassium magazine
(now Fertilizer International)
He is a Fellow of AIChE, a registered Professional Engineer, and past
Chairman of the Central Florida AIChE Section.
#

John Wing, P.E., 1121 Waters Edge Drive, Lakeland FL 33801, USA
Phone: 1-863-666-6555
Email: johnwingpe@verizon.net

23